From alex at chiefmall.com  Wed Feb  3 13:41:36 2010
From: alex at chiefmall.com (Alex Ksikes)
Date: Wed, 03 Feb 2010 04:41:36 -0800
Subject: [Image-SIG] faceted search interface
Message-ID: <E1NceY4-0003cw-KG@mail.chiefmall.com>

Hi all,

I have just put together a system to automatically build a faceted search interface and search engine from any data. Would you mind having a look at it and letting me know what you think of it?

Here is a particular instance on IMDb (the internet movie database). You can refine by particular categories such as genres, actors or directors with the results ordered by user ratings and number of votes.

http://imdb.cloudmining.net/search?query=(@actor%20Brad%20Pitt)

This IMDb instance was built automatically. All that was needed was to setup the data in a specific format and to optionally customize the look and feel of a search result. In fact the system was also applied to DBLP:

http://dblp.cloudmining.net/search?query=(@venue%20CHI)

The system is called Cloud Mining (http://www.cloudmining.net). Your feedback would be very much appreciated.

Thank you so much,

Alex Ksikes

From playfight at mac.com  Mon Feb  8 05:53:01 2010
From: playfight at mac.com (Doug Easterly)
Date: Mon, 08 Feb 2010 17:53:01 +1300
Subject: [Image-SIG] _imaging C module
Message-ID: <B7259DF8-7B8E-4E15-9530-CB3498E1AFA4@mac.com>

Hello

I have a MacPowerbook Pro, Intel Core Duo running 10.6.2
and Python 2.6.1 installed 

I have also tried these different instructions for getting PIL to work:
	http://jetfar.com/libjpeg-and-python-imaging-pil-on-snow-leopard/
	http://gknauth.blogspot.com/2009/10/building-pil-116-on-mac-os-x-106.html
	http://proteus-tech.com/blog/cwt/install-pil-in-snow-leopard/

And I have done all manner of variation on the above (installing on top, deleting everything and working from scratch, using macports, etc).

But, each time I try this simple example: 
>>>	from PIL import Image
>>>	mg = Image.open('test.jpg')
>>>	print(mg.size)
>>>	mg.show()

Everything works until I get to mg.show(). Then I get the following traceback:

File "<stdin>", line 1, in <module>
  File "/Library/Python/2.6/site-packages/PIL/Image.py", line 1449, in show
    _showxv(self, title, command)
  File "/Library/Python/2.6/site-packages/PIL/Image.py", line 2082, in _showxv
    file = image._dump(format=format)
  File "/Library/Python/2.6/site-packages/PIL/Image.py", line 476, in _dump
    self.load()
  File "/Library/Python/2.6/site-packages/PIL/ImageFile.py", line 155, in load
    self.load_prepare()
  File "/Library/Python/2.6/site-packages/PIL/ImageFile.py", line 223, in load_prepare
    self.im = Image.core.new(self.mode, self.size)
  File "/Library/Python/2.6/site-packages/PIL/Image.py", line 36, in __getattr__
    raise ImportError("The _imaging C module is not installed")

Can anyone please help with this? I've been trying for several days now and just seem to be spinning my wheels.

More info: my build seemed to work ok, but my self test had a failure:

PIL 1.1.6 BUILD SUMMARY
--------------------------------------------------------------------
version       1.1.6
platform      darwin 2.6.1 (r261:67515, Jul  7 2009, 23:51:51)
              [GCC 4.2.1 (Apple Inc. build 5646)]
--------------------------------------------------------------------
--- TKINTER support ok
--- JPEG support ok
--- ZLIB (PNG/ZIP) support ok
--- FREETYPE2 support ok

*****************************************************************
Failure in example: _info(Image.open("Images/lena.jpg"))
from line #24 of selftest.testimage
Exception raised:
Traceback (most recent call last):
  File "./doctest.py", line 499, in _run_examples_inner
    exec compile(source, "<string>", "single") in globs
  File "<string>", line 1, in <module>
  File "./selftest.py", line 22, in _info
    im.load()
  File "PIL/ImageFile.py", line 207, in load
    raise IOError(error + " when reading image file")
IOError: decoding error when reading image file
1 items had failures:
   1 of  57 in selftest.testimage
***Test Failed*** 1 failures.
*** 1 tests of 57 failed.



cheers
Doug



From ariel.goldstien at gmail.com  Wed Feb  3 10:06:17 2010
From: ariel.goldstien at gmail.com (ariel goldstien)
Date: Wed, 3 Feb 2010 11:06:17 +0200
Subject: [Image-SIG] proble loading font
Message-ID: <258791b31002030106y1ba605b2t8b6e891df46ad99f@mail.gmail.com>

Hi.
My name is Ariel Goldstien I work at The Hebrew university laboratory.
I am trying to write a Hebrew string to a BMP. and I keep getting "The
_imaging C module is not installed" despite the fact that I have managed to
import _imaging.
this is the code:
import Image
import ImageDraw
import ImageFont
import _imaging

a=u'1?????' #u'..' converts to Unicode
font = ImageFont.truetype('c:/mriam.ttf',24) #standard Hebrew font, this
line throws the error
im=Image.new('RGB',(200,200),(100,100,100))
d=ImageDraw.Draw(im)
d.text((0,0),a)
im.show()
I have spent more than 3 dayes trying to figure it out. any help would be
very appriciated
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100203/510fb409/attachment-0001.htm>

From espen at medialog.no  Wed Feb 10 12:01:24 2010
From: espen at medialog.no (Espen Moe-Nilssen)
Date: Wed, 10 Feb 2010 12:01:24 +0100
Subject: [Image-SIG] Colorize Image
Message-ID: <BA4D0AEA-0B2C-4E7D-90A2-7708A967376D@medialog.no>

Hi

We are making an open source project, http://plone.org/products/subskins
I have been trying over and over again to find a way to let PIL  
colorize an image.

It should work like this:
1) I have a greyscale image
2) A color is selected, for example "#123456"
3) The image is colorized in this way:
     Whatever was white is still white
     Whatever was black (100%) is now "#123456"
     Whaterver was grey (50%) is now 50% of "#123456"

I have tried a lot of different approaches, but the colors always  
come out too dark. Is there a way to do this ?

The code is here:
http://svn.plone.org/svn/collective/Products.PloneSubSkins/trunk/ 
Products.PloneSubSkins/Products/PloneSubSkins/browser/views.py

where the lines about this are:

______________________________________


class ColorizedImage(BrowserView):
     def __call__(self):
         portal_skins = getToolByName(self.context, 'portal_skins')
         img = portal_skins.restrictedTraverse(self.request.img)
         input = StringIO()
         input.write(img._data)
         input.seek(0)
         image = Image.open(input)
         newimage = colorize(image, self.request.color)
         output = StringIO()
         newimage.save(output, format='PNG')
         self.request.response.setHeader('Content-Type','image/png')
         return output.getvalue()

def colorize(image, color):
     color = color[1:]
     assert len(color) == 6 or len(color)==3
     if len(color)==6:
         r,g,b = int(color[:2],16),int(color[2:4],16),int(color[4:7],16)
     elif len(color)==3:
         r,g,b = int(color[0],16)*16,int(color[1],16)*16,int(color[2], 
16)*16
     palette = make_linear_ramp((r,g,b))
     if image.mode != "L":
         image = image.convert('L')
     image.putpalette(palette)
     return image

#Thanks to http://effbot.org/zone/pil-sepia.htm for the snippet
def make_linear_ramp(white):
     # putpalette expects [r,g,b,r,g,b,...]
     ramp = []
     r, g, b = white
     for i in range(255):
         ramp.extend((r*i/255, g*i/255, b*i/255))
     return ramp

__________________________________________


Espen










-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100210/e83e7a1d/attachment-0001.htm>

From jcupitt at gmail.com  Thu Feb 11 12:50:11 2010
From: jcupitt at gmail.com (jcupitt at gmail.com)
Date: Thu, 11 Feb 2010 11:50:11 +0000
Subject: [Image-SIG] Colorize Image
In-Reply-To: <BA4D0AEA-0B2C-4E7D-90A2-7708A967376D@medialog.no>
References: <BA4D0AEA-0B2C-4E7D-90A2-7708A967376D@medialog.no>
Message-ID: <522c6461002110350p6208e7a7ic41820b88dc75de@mail.gmail.com>

On 10 February 2010 11:01, Espen Moe-Nilssen <espen at medialog.no> wrote:
> We are making an open source project,?http://plone.org/products/subskins
> I have been trying over and over again to find a way to let PIL colorize an
> image.
> It should work like this:
> 1) I have a greyscale image
> 2) A color is selected, for example "#123456"
> 3) The image is colorized in this way:
> ?? ?Whatever was white is still white
> ?? ?Whatever was black (100%) is now "#123456"
> ?? ?Whaterver was grey (50%) is now 50% of "#123456"
> I have tried a lot of different approaches, but the colors always come out
> too dark.?Is there a way to do this ?

You need to split the image into luminance (your greyscale image) and
colour (the constant colour you want to apply).

If you imagine the RGB space as a cube, the neutral axis (all the grey
colours) is a straight line from (0, 0, 0) to (255, 255, 255). What
you want to do is translate that line (ie. keep the angle of the thing
the same, just translate the whole line up/down/left/right/fwd/back)
so that it passes through #123456 instead. Any bits that go over 255
or under 0 need to be clipped.

This is easy in CIELAB space, but still OK in RGB.

John

From spe.stani.be at gmail.com  Thu Feb 11 20:26:15 2010
From: spe.stani.be at gmail.com (Stani)
Date: Thu, 11 Feb 2010 20:26:15 +0100
Subject: [Image-SIG] PIL not saving metadata from progressive Jpeg's
In-Reply-To: <d547c431001270550o61f1a2f1r7cb0ebde2d874257@mail.gmail.com>
References: <d547c431001270550o61f1a2f1r7cb0ebde2d874257@mail.gmail.com>
Message-ID: <2078a7ad1002111126g65623ccbjed87a75be869cffc@mail.gmail.com>

http://tilloy.net/dev/pyexiv2/index.htm

On Wed, Jan 27, 2010 at 2:50 PM, Frank Borell <frankborell at gmail.com> wrote:
> Hello,
>
> I've tested on PIL 1.16 and 1.17. Metadata(IPTC and EXIF) are saved
> fine when the jpg image is NOT progressive. But when the jpg is
> progressive, the metadata gets wiped out. Does anyone have any ideas
> on how to preserve?
>
> im = Image.open('progressive jpeg path')
>
> im.save('jpeg path') or
> im.save('jpeg path',option={'progression':False,'quality':
> 60,'optimize':True})
>
> Thanks,
> Frank
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>



-- 
Phatch Photo Batch Processor - http://photobatch.stani.be
SPE Python IDE - http://pythonide.stani.be

From spe.stani.be at gmail.com  Thu Feb 11 20:29:39 2010
From: spe.stani.be at gmail.com (Stani)
Date: Thu, 11 Feb 2010 20:29:39 +0100
Subject: [Image-SIG] Colorize Image
In-Reply-To: <522c6461002110350p6208e7a7ic41820b88dc75de@mail.gmail.com>
References: <BA4D0AEA-0B2C-4E7D-90A2-7708A967376D@medialog.no>
	<522c6461002110350p6208e7a7ic41820b88dc75de@mail.gmail.com>
Message-ID: <2078a7ad1002111129i1a3b4caq1e581c6088ccdf1e@mail.gmail.com>

How about using ImageOps.colorize ;-) ?

On Thu, Feb 11, 2010 at 12:50 PM,  <jcupitt at gmail.com> wrote:
> On 10 February 2010 11:01, Espen Moe-Nilssen <espen at medialog.no> wrote:
>> We are making an open source project,?http://plone.org/products/subskins
>> I have been trying over and over again to find a way to let PIL colorize an
>> image.
>> It should work like this:
>> 1) I have a greyscale image
>> 2) A color is selected, for example "#123456"
>> 3) The image is colorized in this way:
>> ?? ?Whatever was white is still white
>> ?? ?Whatever was black (100%) is now "#123456"
>> ?? ?Whaterver was grey (50%) is now 50% of "#123456"
>> I have tried a lot of different approaches, but the colors always come out
>> too dark.?Is there a way to do this ?
>
> You need to split the image into luminance (your greyscale image) and
> colour (the constant colour you want to apply).
>
> If you imagine the RGB space as a cube, the neutral axis (all the grey
> colours) is a straight line from (0, 0, 0) to (255, 255, 255). What
> you want to do is translate that line (ie. keep the angle of the thing
> the same, just translate the whole line up/down/left/right/fwd/back)
> so that it passes through #123456 instead. Any bits that go over 255
> or under 0 need to be clipped.
>
> This is easy in CIELAB space, but still OK in RGB.
>
> John
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>



-- 
Phatch Photo Batch Processor - http://photobatch.stani.be
SPE Python IDE - http://pythonide.stani.be

From cannon.el at gmail.com  Thu Feb 11 17:12:25 2010
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Thu, 11 Feb 2010 08:12:25 -0800
Subject: [Image-SIG] Colorize Image
In-Reply-To: <522c6461002110350p6208e7a7ic41820b88dc75de@mail.gmail.com>
References: <BA4D0AEA-0B2C-4E7D-90A2-7708A967376D@medialog.no>
	<522c6461002110350p6208e7a7ic41820b88dc75de@mail.gmail.com>
Message-ID: <b986b291002110812y6af8d104t48f86008bec79026@mail.gmail.com>

There is a function for this. from the handbook:
ImageOps.colorize(image, black, white) => image

Colorize grayscale image. The black and white arguments should be RGB
tuples or color names; this function calculates a colour wedge mapping
all black pixels in the source image to the first colour, and all
white pixels to the second colour.

call it with your color for black and white for white, it accepts all
"standard" color specifications, such as "#ff2253" or (255, 0, 0).

On Thu, Feb 11, 2010 at 3:50 AM,  <jcupitt at gmail.com> wrote:
> On 10 February 2010 11:01, Espen Moe-Nilssen <espen at medialog.no> wrote:
>> We are making an open source project,?http://plone.org/products/subskins
>> I have been trying over and over again to find a way to let PIL colorize an
>> image.
>> It should work like this:
>> 1) I have a greyscale image
>> 2) A color is selected, for example "#123456"
>> 3) The image is colorized in this way:
>> ?? ?Whatever was white is still white
>> ?? ?Whatever was black (100%) is now "#123456"
>> ?? ?Whaterver was grey (50%) is now 50% of "#123456"
>> I have tried a lot of different approaches, but the colors always come out
>> too dark.?Is there a way to do this ?
>
> You need to split the image into luminance (your greyscale image) and
> colour (the constant colour you want to apply).
>
> If you imagine the RGB space as a cube, the neutral axis (all the grey
> colours) is a straight line from (0, 0, 0) to (255, 255, 255). What
> you want to do is translate that line (ie. keep the angle of the thing
> the same, just translate the whole line up/down/left/right/fwd/back)
> so that it passes through #123456 instead. Any bits that go over 255
> or under 0 need to be clipped.
>
> This is easy in CIELAB space, but still OK in RGB.
>
> John
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>

From mathewsl at indiana.edu  Thu Feb 11 22:51:46 2010
From: mathewsl at indiana.edu (Mathews, Leah Nicole)
Date: Thu, 11 Feb 2010 16:51:46 -0500
Subject: [Image-SIG] compatibility
Message-ID: <3B184B0C-3225-4A14-8EB5-E18CAE953007@indiana.edu>

Hello,

> We are currently using Python Imaging Library 1.1.6 and evaluating 
> for use with snow leopard. Is it compatible and 
> what is the version that is compatible?
> 
> Thanks,
> Leah

From espen at medialog.no  Fri Feb 12 00:29:09 2010
From: espen at medialog.no (Espen Moe-Nilssen)
Date: Fri, 12 Feb 2010 00:29:09 +0100
Subject: [Image-SIG] Colorize Image
In-Reply-To: <b986b291002110812y6af8d104t48f86008bec79026@mail.gmail.com>
References: <BA4D0AEA-0B2C-4E7D-90A2-7708A967376D@medialog.no>
	<522c6461002110350p6208e7a7ic41820b88dc75de@mail.gmail.com>
	<b986b291002110812y6af8d104t48f86008bec79026@mail.gmail.com>
Message-ID: <E1NfiT7-0004Rz-8j@webmail02.domeneshop.no>


Thanks a lot. I thought there had to be one, but I never found it.
This will come in very handy

Espen,

>There is a function for this. from the handbook:
>ImageOps.colorize(image, black, white) => image
>
>Colorize grayscale image. The black and white arguments should be RGB
>tuples or color names; this function calculates a colour wedge mapping
>all black pixels in the source image to the first colour, and all
>white pixels to the second colour.
>
>call it with your color for black and white for white, it accepts all
>"standard" color specifications, such as "#ff2253" or (255, 0, 0).
>
>On Thu, Feb 11, 2010 at 3:50 AM,  <jcupitt at gmail.com> wrote:
>> On 10 February 2010 11:01, Espen Moe-Nilssen <espen at medialog.no> wrote:
>>> We are making an open source project,?http://plone.org/products/subskins
>>> I have been trying over and over again to find a way to let PIL colorize an
>>> image.
>>> It should work like this:
>>> 1) I have a greyscale image
>>> 2) A color is selected, for example "#123456"
>>> 3) The image is colorized in this way:
>>> ?? ?Whatever was white is still white
>>> ?? ?Whatever was black (100%) is now "#123456"
>>> ?? ?Whaterver was grey (50%) is now 50% of "#123456"
>>> I have tried a lot of different approaches, but the colors always come out
>>> too dark.?Is there a way to do this ?
>>
>> You need to split the image into luminance (your greyscale image) and
>> colour (the constant colour you want to apply).
>>
>> If you imagine the RGB space as a cube, the neutral axis (all the grey
>> colours) is a straight line from (0, 0, 0) to (255, 255, 255). What
>> you want to do is translate that line (ie. keep the angle of the thing
>> the same, just translate the whole line up/down/left/right/fwd/back)
>> so that it passes through #123456 instead. Any bits that go over 255
>> or under 0 need to be clipped.
>>
>> This is easy in CIELAB space, but still OK in RGB.
>>
>> John
>> _______________________________________________
>> Image-SIG maillist ?- ?Image-SIG at python.org
>> http://mail.python.org/mailman/listinfo/image-sig
>>


-- 
Espen Moe-Nilssen
Grieg Medialog AS

From fredrik at pythonware.com  Thu Feb 11 21:06:18 2010
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 11 Feb 2010 21:06:18 +0100
Subject: [Image-SIG] delayed mail
Message-ID: <368a5cd51002111206w543da989r9031737b3d0631da@mail.gmail.com>

sorry for the delays in mail handling; I've been on the road more or
less constantly since late november, and more mails than usual have
ended up in the moderation queues lately (most of it is indeed spam,
but not everything...).

</F>

From fredrik at pythonware.com  Fri Feb 12 03:01:16 2010
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 12 Feb 2010 03:01:16 +0100
Subject: [Image-SIG] delayed mail
In-Reply-To: <368a5cd51002111206w543da989r9031737b3d0631da@mail.gmail.com>
References: <368a5cd51002111206w543da989r9031737b3d0631da@mail.gmail.com>
Message-ID: <368a5cd51002111801w34ec9a8x8ff9f92010dc82b5@mail.gmail.com>

(and of course that mail got caught by the filter as well.  argh!)

On Thu, Feb 11, 2010 at 9:06 PM, Fredrik Lundh <fredrik at pythonware.com> wrote:
> sorry for the delays in mail handling; I've been on the road more or
> less constantly since late november, and more mails than usual have
> ended up in the moderation queues lately (most of it is indeed spam,
> but not everything...).
>
> </F>
>

From fredrik at pythonware.com  Fri Feb 12 03:55:03 2010
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 12 Feb 2010 03:55:03 +0100
Subject: [Image-SIG] compatibility
In-Reply-To: <3B184B0C-3225-4A14-8EB5-E18CAE953007@indiana.edu>
References: <3B184B0C-3225-4A14-8EB5-E18CAE953007@indiana.edu>
Message-ID: <368a5cd51002111855h5a8397dfp72b65a93dd60e889@mail.gmail.com>

On Thu, Feb 11, 2010 at 10:51 PM, Mathews, Leah Nicole
<mathewsl at indiana.edu> wrote:
> Hello,
>
>> We are currently using Python Imaging Library 1.1.6 and evaluating
>> for use with snow leopard. Is it compatible and
>> what is the version that is compatible?

I'm not aware of any known incompatibilities with any version of Mac
OS X, for any version of PIL.  You need a version of PIL that matches
your Python installation, but that's about it.

</F>

From fredrik at pythonware.com  Fri Feb 12 03:59:26 2010
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 12 Feb 2010 03:59:26 +0100
Subject: [Image-SIG] How to pass the address of the buffer to
	Image.frombuffer()
In-Reply-To: <fb3ed3f71001260351n570b7a47p5353ee99849840ba@mail.gmail.com>
References: <fb3ed3f71001260351n570b7a47p5353ee99849840ba@mail.gmail.com>
Message-ID: <368a5cd51002111859v148eaab2of4d034935cd5e986@mail.gmail.com>

2010/1/26 ??????? <mail at vpetkov.com>:
> I want to display captured frames from a webcam. I am using
> video4linux2, v4l2 bindings for python and ctypes.
>
> I am also using Tk. In the main loop of my v4l2 app, after Q_BUF,
> STREAM_ON, DQ_BUF I call method that looks something like that:
>
>
> window = Tkinter.Tk()
> image = ImageTk.PhotoImage(pixelfmt, (320, 240))
> label = Tkinter.Label(window, image=image)
> label.pack()
>
> im = Image.frombuffer(pixelfmt, (320, 240),
> ctypes.string_at(buf.userptr, buf.length), 'raw', pixelfmt, 0, 1)
>
> image.paste(im)
> window.update()
>
> Now, how is buf.userptr supposed to look like? A ctypes.c_void_p?
> In my case buf.userptr is long. Something like 170692608.
>
> I don't know how addresses in python should look like.

Did you figure this out?  I think this is more of a ctypes question
than a PIL question, so you might want to repost this to a more
general Python forum, such as python-list at python.org (but feel free
to follow up here with the solution if you have it).

Cheers /F

From fredrik at pythonware.com  Fri Feb 12 08:42:02 2010
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 12 Feb 2010 08:42:02 +0100
Subject: [Image-SIG] proble loading font
In-Reply-To: <258791b31002030106y1ba605b2t8b6e891df46ad99f@mail.gmail.com>
References: <258791b31002030106y1ba605b2t8b6e891df46ad99f@mail.gmail.com>
Message-ID: <368a5cd51002112342p72c7f3c3p46999cc946ab5114@mail.gmail.com>

What platform is this?  If you start a Python interpreter and write:

    >>> import Image
    >>> Image.core

what does it output?

</F>

On Wed, Feb 3, 2010 at 10:06 AM, ariel goldstien
<ariel.goldstien at gmail.com> wrote:
> Hi.
> My name is Ariel Goldstien I work at The Hebrew university laboratory.
> I am trying to write a Hebrew string to a BMP. and I keep getting "The
> _imaging C module is not installed" despite the fact that I have managed to
> import _imaging.
> this is the code:
> import Image
> import ImageDraw
> import ImageFont
> import _imaging
>
> a=u'1?????' #u'..' converts to Unicode
> font = ImageFont.truetype('c:/mriam.ttf',24) #standard Hebrew font, this
> line throws the error
> im=Image.new('RGB',(200,200),(100,100,100))
> d=ImageDraw.Draw(im)
> d.text((0,0),a)
> im.show()
> I have spent more than 3 dayes trying to figure it out. any help would be
> very appriciated
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>

From Chris.Barker at noaa.gov  Fri Feb 12 18:28:35 2010
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Fri, 12 Feb 2010 09:28:35 -0800
Subject: [Image-SIG] Is PIL 1.1.7 final "official"
In-Reply-To: <368a5cd50912111646x9ff190cib0da871c50807d2f@mail.gmail.com>
References: <800886C2A4A73E44BA0FE45152C023705D78630E8C@ADSK-NAMSG-02.MGDADSK.autodesk.com>
	<368a5cd50912111646x9ff190cib0da871c50807d2f@mail.gmail.com>
Message-ID: <4B758FC3.10401@noaa.gov>

Fredrik Lundh wrote:
> Yeah, 1.1.7 is official; it's just me having accidentally skipped a
> couple of steps in the release procedure.

Do you have any Mac binaries? If not, do you have anyone to build them? 
If not, should I try to get it done?

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R            (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From nicolas.roux at st.com  Fri Feb 12 10:35:16 2010
From: nicolas.roux at st.com (Nicolas ROUX)
Date: Fri, 12 Feb 2010 10:35:16 +0100
Subject: [Image-SIG] Support of RGB 16 bits images
Message-ID: <000001caabc6$af860270$d4e8810a@gnb.st.com>

Hello,
We need to load/save RGB images in more than 8 bits and up to 16bits.
We use to work with TIFF, PNG, BMP, ...

Is there a plan to support such RGB bitwidth (and so associated file format) ?

Thanks,

Cheers,
Nicolas.


From cannon.el at gmail.com  Sat Feb 13 00:05:01 2010
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Fri, 12 Feb 2010 15:05:01 -0800
Subject: [Image-SIG] BUG? PIL does not recognise alpha channel in
	attached png file
In-Reply-To: <4B5E748F.2090902@smithpolglase.com>
References: <4B5E748F.2090902@smithpolglase.com>
Message-ID: <b986b291002121505y58f37bf2j7323c0b2dc48d8f4@mail.gmail.com>

I confirm the results PIL 1.16 Python 2.5.2
Seems pretty odd. What program did you use to create the image?

On Mon, Jan 25, 2010 at 8:50 PM, Root <root at smithpolglase.com> wrote:
> Hi,
>
> The attached image shows an alpha channel when viewed with gimp or eog, but
> not when loading with PIL (version 1.1.6-3ubuntu1 as supplied with Ubuntu
> Karmic AMD64 Desktop):
>
>
>>>> import Image
>>>> im = Image.open("seabreeze.png")
>>>> print im.format
> PNG
>>>> print im.size
> (599, 264)
>>>> print im.mode
> RGB
>
>
> When I save the file using im.save(), and then open with gimp or eog, the
> alpha channel is removed (to be expected given that the mode is read as
> RGB).
>
> Is this is a bug?
>
>
> Cheers,
> Justin.
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>

From justin at smithpolglase.com  Sat Feb 13 05:49:40 2010
From: justin at smithpolglase.com (Justin Smith)
Date: Sat, 13 Feb 2010 12:49:40 +0800
Subject: [Image-SIG] BUG? PIL does not recognise alpha channel in
 attached png file
In-Reply-To: <b986b291002121505y58f37bf2j7323c0b2dc48d8f4@mail.gmail.com>
References: <4B5E748F.2090902@smithpolglase.com>
	<b986b291002121505y58f37bf2j7323c0b2dc48d8f4@mail.gmail.com>
Message-ID: <4B762F64.7080008@smithpolglase.com>

Laura & Edward Cannon wrote:
> I confirm the results PIL 1.16 Python 2.5.2
> Seems pretty odd. What program did you use to create the image?

I didn't create the image - it is generated by a weather web-site. I 
have a python script that downloads it for use on my MythTV box.

BTW I'm using Python 2.6.4.


Cheers,
Justin.

From cannon.el at gmail.com  Mon Feb 15 01:26:52 2010
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Sun, 14 Feb 2010 16:26:52 -0800
Subject: [Image-SIG] Support of RGB 16 bits images
In-Reply-To: <000001caabc6$af860270$d4e8810a@gnb.st.com>
References: <000001caabc6$af860270$d4e8810a@gnb.st.com>
Message-ID: <b986b291002141626o49435236pc89a106adf2118d@mail.gmail.com>

I have a similar question. PIL supports 16 bit integer for single band
images, but not multi-band images. Messing around a bit with the 1.16
version, I find that many of the operations in the library are for 8
bit bands only. ImageMagick has python bindings and supports a wide
variety of formats, check it out.

On Fri, Feb 12, 2010 at 1:35 AM, Nicolas ROUX <nicolas.roux at st.com> wrote:
> Hello,
> We need to load/save RGB images in more than 8 bits and up to 16bits.
> We use to work with TIFF, PNG, BMP, ...
>
> Is there a plan to support such RGB bitwidth (and so associated file format) ?
>
> Thanks,
>
> Cheers,
> Nicolas.
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>

From cannon.el at gmail.com  Wed Feb 17 06:57:18 2010
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Tue, 16 Feb 2010 21:57:18 -0800
Subject: [Image-SIG] ImageStat silent convert
Message-ID: <b986b291002162157w67545925u4f78b9e1bfb0a434@mail.gmail.com>

while using PIL 1.16 recently, I noticed that ImageStat seems to do a
silent convert to mode L when given a mode F image. Looping though the
image manually using the im.load() access object gives me a different
maximum than ImageStat.Stat(im).extrema. Is this intentional? it
doesn't seem to be documented. Is it perhaps been fixed?
Edward
Unicorn School

From wrybread at gmail.com  Wed Feb 17 12:30:23 2010
From: wrybread at gmail.com (Alec Bennett)
Date: Wed, 17 Feb 2010 03:30:23 -0800
Subject: [Image-SIG] How to improve rotation quality?
Message-ID: <cab7082b1002170330i6ee23269y79fbc2a235dae932@mail.gmail.com>

I'm wondering if anyone has any idea how to get better results from PIL's
rotate() function? No matter what filter I use I'm getting very jagged edges
after rotating an image.

Here's how I'm invoking the filters:

pic = pic.rotate(random_rotation, resample=Image.NEAREST, expand=1)
pic = pic.rotate(random_rotation, resample=Image.BILINEAR, expand=1)
pic = pic.rotate(random_rotation, resample=Image.BICUBIC, expand=1)    # the
best I think

I posted a sample app that opens an image and rotates it using each of the
filters. You can see the jagged edges pretty plainly.

I have a slideshow app that I'm working on where these edges are very
obvious, so I'm wondering if anyone might have any tips on getting better
results from rotate(), or from some other method or module that might do a
better job?

My very simple sample app that tests each of the filters is here:

http://sinkingsensation.com/stuff/rotation_test.zip
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100217/eca5d534/attachment.htm>

From cannon.el at gmail.com  Wed Feb 17 23:12:41 2010
From: cannon.el at gmail.com (Laura & Edward Cannon)
Date: Wed, 17 Feb 2010 14:12:41 -0800
Subject: [Image-SIG] How to improve rotation quality?
In-Reply-To: <cab7082b1002170330i6ee23269y79fbc2a235dae932@mail.gmail.com>
References: <cab7082b1002170330i6ee23269y79fbc2a235dae932@mail.gmail.com>
Message-ID: <b986b291002171412l4c999348k1f82d2211b6d8ee@mail.gmail.com>

Try using Image.ANTIALIAS, I find it has the best quality.

On Wed, Feb 17, 2010 at 3:30 AM, Alec Bennett <wrybread at gmail.com> wrote:
> I'm wondering if anyone has any idea how to get better results from PIL's
> rotate() function? No matter what filter I use I'm getting very jagged edges
> after rotating an image.
>
> Here's how I'm invoking the filters:
>
> pic = pic.rotate(random_rotation, resample=Image.NEAREST, expand=1)
> pic = pic.rotate(random_rotation, resample=Image.BILINEAR, expand=1)
> pic = pic.rotate(random_rotation, resample=Image.BICUBIC, expand=1)??? # the
> best I think
>
> I posted a sample app that opens an image and rotates it using each of the
> filters. You can see the jagged edges pretty plainly.
>
> I have a slideshow app that I'm working on where these edges are very
> obvious, so I'm wondering if anyone might have any tips on getting better
> results from rotate(), or from some other method or module that might do a
> better job?
>
> My very simple sample app that tests each of the filters is here:
>
> http://sinkingsensation.com/stuff/rotation_test.zip
>
>
>
> _______________________________________________
> Image-SIG maillist ?- ?Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>

From Chris.Barker at noaa.gov  Wed Feb 17 23:38:03 2010
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Wed, 17 Feb 2010 14:38:03 -0800
Subject: [Image-SIG] How to improve rotation quality?
In-Reply-To: <cab7082b1002170330i6ee23269y79fbc2a235dae932@mail.gmail.com>
References: <cab7082b1002170330i6ee23269y79fbc2a235dae932@mail.gmail.com>
Message-ID: <4B7C6FCB.2000909@noaa.gov>

Alec Bennett wrote:
> I'm wondering if anyone has any idea how to get better results from 
> PIL's rotate() function? No matter what filter I use I'm getting very 
> jagged edges after rotating an image.
> 
> Here's how I'm invoking the filters:
> 
> pic = pic.rotate(random_rotation, resample=Image.NEAREST, expand=1)   
> pic = pic.rotate(random_rotation, resample=Image.BILINEAR, expand=1)   
> pic = pic.rotate(random_rotation, resample=Image.BICUBIC, expand=1)    # 
> the best I think

Too bad ANTIALIAS doesn't seem to be available for rotate -- darn. It 
might look better.

But to some extent, you can only get so good rotating an image.

However, I took at look at your test images -- it looked like better 
smoothing was going on inside the white border than outside, which made 
me think -- when rotating, if you are going to use interpolation, how 
does it interpolate to outside the image? It doesn't. So I tried adding 
a black background to the image first, then rotating it -- much better. 
See the enclosed version of your test code, and a rotated image.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R            (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.py
Type: application/x-python
Size: 961 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100217/b9ba0d63/attachment-0001.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test-3.jpg
Type: image/jpeg
Size: 42962 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100217/b9ba0d63/attachment-0001.jpg>

From johan at slentrian.org  Thu Feb 18 10:23:48 2010
From: johan at slentrian.org (Johan Forsberg)
Date: Thu, 18 Feb 2010 10:23:48 +0100
Subject: [Image-SIG] Floodfill
Message-ID: <27d300711002180123p53644bb5pe35899c6743529cf@mail.gmail.com>

Hello,

I have been using PIL to write a simple drawing program and it's been great
fun. The only thing I've really been missing from the library is a fast
flood-fill routine. The included python based floodfill, while functional,
is very slow (in 1.1.7), I guess due to the per-pixel access required. Since
I want interactive speed, I decided to try and hack in a floodfill function
in C. Instead of reinventing the wheel, I googled up a nice implementation
in the "graphics gems" collection (
http://tog.acm.org/resources/GraphicsGems/) which needed only small
modifications in order to work with the PIL code. I havent benchmarked it
but it's (not surprisingly) orders of magnitudes faster than the python
routine. I've only implemented it for 8-bit palette image formats since
that's what I use, but it should be easy to generalize.

Would it be a good idea to clean this up and submit it as a patch to PIL?
The graphics gems license is very liberal and seems to me to be compatible
(this is what I found on the webpage above):

---8<---
The Graphics Gems code is copyright-protected. In other words, you cannot
claim the text of the code as your own and resell it. Using the code is
permitted in any program, product, or library, non-commercial or commercial.
Giving credit is not required, though is a nice gesture.
---8<---

Cheers,
Johan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100218/88a045b0/attachment.htm>

From wrybread at gmail.com  Thu Feb 18 12:50:01 2010
From: wrybread at gmail.com (Alec Bennett)
Date: Thu, 18 Feb 2010 03:50:01 -0800
Subject: [Image-SIG] How to improve rotation quality?
In-Reply-To: <4B7C6FCB.2000909@noaa.gov>
References: <cab7082b1002170330i6ee23269y79fbc2a235dae932@mail.gmail.com>
	<4B7C6FCB.2000909@noaa.gov>
Message-ID: <cab7082b1002180350i2f0c115fj8289e4244fbe57@mail.gmail.com>

> So I tried adding a black background to the image first, then rotating it
-- much better.
> See the enclosed version of your test code, and a rotated image.

Wow, what a difference! Even if the border is a single pixel, much better
results. I set the border color to grey, which seems to work as well as
black, and am now a happy slideshow maker.

Huge thanks.

And to anyone else coming down this road, if the resampling filters aren't
having any effect, update to the latest PIL.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100218/a9849d9d/attachment.htm>

From karsten.hiddemann at mathematik.uni-dortmund.de  Fri Feb 19 21:53:25 2010
From: karsten.hiddemann at mathematik.uni-dortmund.de (Karsten Hiddemann)
Date: Fri, 19 Feb 2010 21:53:25 +0100
Subject: [Image-SIG] Floodfill
In-Reply-To: <27d300711002180123p53644bb5pe35899c6743529cf@mail.gmail.com>
References: <27d300711002180123p53644bb5pe35899c6743529cf@mail.gmail.com>
Message-ID: <4B7EFA45.5020805@mathematik.uni-dortmund.de>

Johan Forsberg wrote:
> Hello,
> 
> I have been using PIL to write a simple drawing program and it's been 
> great fun. The only thing I've really been missing from the library is a 
> fast flood-fill routine. The included python based floodfill, while 
> functional, is very slow (in 1.1.7), I guess due to the per-pixel access 
> required. Since I want interactive speed, I decided to try and hack in a 
> floodfill function in C.

Just wanted to add that I remember that somebody else did that as well. 
This is a post from the list by Douglas Bagnall on 01.06.2009 02:23, 
original title "[Image-SIG] remove an image background using PIL"

> Peter Yen <colorpyen at gmail.com> wrote:
> 
>> Thanks for your quick response. The background is unknown at the time of
>> processing. Actually I don't need a very accurate methodology to remove all
>> background, removing partially is good enough.
> 
> I once wrote a (GPL'd) C module that fairly reliably extracts the
> background of a photograph of a drawing on paper.  It is essentially the
> same as PIL's floodfill but it follows shallow gradients.
> 
> The public git browser[1] seems to be broken right now, but if this
> sounds like what you need download a tarball[2], and look for the
> expand_region function in the img-c directory.
> 
> [1] https://savannah.nongnu.org/git/?group=tetuhi
> [2] http://sourceforge.net/project/showfiles.php?group_id=217385
> 
> 
> Douglas

Cheers,
Karsten

From Chris.Barker at noaa.gov  Fri Feb 19 23:42:57 2010
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Fri, 19 Feb 2010 14:42:57 -0800
Subject: [Image-SIG] How to improve rotation quality?
In-Reply-To: <cab7082b1002180350i2f0c115fj8289e4244fbe57@mail.gmail.com>
References: <cab7082b1002170330i6ee23269y79fbc2a235dae932@mail.gmail.com>
	<4B7C6FCB.2000909@noaa.gov>
	<cab7082b1002180350i2f0c115fj8289e4244fbe57@mail.gmail.com>
Message-ID: <4B7F13F1.9020200@noaa.gov>

Alec Bennett wrote:
>  > So I tried adding a black background to the image first, then 
> rotating it -- much better.
>  > See the enclosed version of your test code, and a rotated image.
> 
> Wow, what a difference! Even if the border is a single pixel, much 
> better results. I set the border color to grey, which seems to work as 
> well as black, and am now a happy slideshow maker.

Ideally, you should use whatever color is going to be in the background 
when you show the image.

Even better would be if you could blend it with a transparent 
background, and show it that way, but that would require an alpha 
channel, and I don't think jpeg supports that.

Glad it worked for you,

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R            (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From uricchio at gmail.com  Mon Feb 22 16:38:08 2010
From: uricchio at gmail.com (Tiberio Uricchio)
Date: Mon, 22 Feb 2010 16:38:08 +0100
Subject: [Image-SIG] Radius bug with GaussianBlur
Message-ID: <b53becb31002220738t5dc3d2f2p6359499f0d7251d8@mail.gmail.com>

Hi,
i just wanted to report a simple bug in ImageFilter.GaussianBlur:

class GaussianBlur(Filter):
    name = "GaussianBlur"

    def __init__(self, radius=2):
        self.radius = 2

    def filter(self, image):
        return image.gaussian_blur(self.radius)

As you can see, radius is always fixed at value 2 instead of using the
argument value.

Tiberio

From wrybread at gmail.com  Sat Feb 20 04:39:42 2010
From: wrybread at gmail.com (Alec Bennett)
Date: Fri, 19 Feb 2010 19:39:42 -0800
Subject: [Image-SIG] How to improve rotation quality?
In-Reply-To: <4B7F13F1.9020200@noaa.gov>
References: <cab7082b1002170330i6ee23269y79fbc2a235dae932@mail.gmail.com>
	<4B7C6FCB.2000909@noaa.gov>
	<cab7082b1002180350i2f0c115fj8289e4244fbe57@mail.gmail.com>
	<4B7F13F1.9020200@noaa.gov>
Message-ID: <cab7082b1002191939t4d8b381od29bd9611281121c@mail.gmail.com>

> Ideally, you should use whatever color is going to be in the background
when you show the image.

Agreed, but the problem is I'm making a collage of pictures where the new
pictures are placed above previous pictures, so there's no one consistent
background color.

> Even better would be if you could blend it with a transparent background,
and show it that way,
> but that would require an alpha channel, and I don't think jpeg supports
that.

I'm converting to BMP and assigning an alpha channel so this is indeed a
possibility. Another thought I had was making a drop shadow effect, which
the thin gray edge outside of my white border already approximates.

Interestingly, when I've tried making a transparent canvas in PIL (which I'd
have to do to add an alpha border) in the past, I consistently got weird
artifacts in the transparent areas. The workaround I used was to make a
small blank PNG that was pure alpha and resize it as necessary.

If I test out the alpha border I'll post back to this thread.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100219/c9cd414d/attachment.html>

From lutz at rmi.net  Sun Feb 21 14:53:22 2010
From: lutz at rmi.net (lutz at rmi.net)
Date: Sun, 21 Feb 2010 08:53:22 -0500 (GMT-05:00)
Subject: [Image-SIG] PIL for Python 3.X?
Message-ID: <6392297.1266760402700.JavaMail.root@mswamui-bichon.atl.sa.earthlink.net>

(Having trouble getting in touch with Fredrik on short notice,
so I'm giving this list a shot).

Does anyone have a definitive answer on if/when PIL will become
available for Python 3.1?  It seems to have stalled last year.
Barring that, are there any reasonable and portable alternatives
for using jpegs in tkinter?  Either way, this needs to be an
official-type release, that can be used by the masses.

I'm updating the book Programming Python for 3.X (only), and 
it uses PIL extensively for image display, thumbnails, and the
like.  Using 2.X-compatible code for the PIL examples isn't
a great option; tkinter's naming/structure has changed much.
And I can't really use PIL at all if its 3.X future is unclear.

Unfortunately, I need a decision on PIL for 3.X in the next 
week or two due to book schedules (and can't work on a port
myself for the same reason).  My last resort is to use one 
of tk's native image formats and relegate PIL to a footnote.

Thanks,
--Mark Lutz
 

From sflist at ihonk.com  Mon Feb 22 08:05:13 2010
From: sflist at ihonk.com (Steve Freitas)
Date: Sun, 21 Feb 2010 23:05:13 -0800
Subject: [Image-SIG] n00b Question
Message-ID: <1266822313.16307.414.camel@phat>

Hi all,

I've downloaded image data[1] which has no header and is comprised of
16-bit signed big-endian pixels (it's a heightmap). I figured I'd cull a
portion of the data and turn it into a TIFF so I could check it out
visually to make sure it looks kinda like a heightmap should. However, I
seem to be doing something wrong. I've read the data into a string (not
unicode, and the file was opened with 'rb'), and I'm using essentially
this code to generate a TIFF:

f = open('data.bin', 'rb')
data = f.read(<a subset of the file representing, say, Colorado>)
p = Image.fromstring('F', (tile_width, tile_height), data, 'raw',
'F;16BS', 0, 1)
p.save('foo.tiff')

I've verified that I'm pulling data that isn't just a bunch of zeros,
and I've verified that the generated TIFF isn't just a bunch of zeros,
but when I view it I get the checkerboard pattern which seems to
indicate a translucent image. Or something.

Part of my confusion comes from the interrelation between the two "mode"
arguments to Image.fromstring(). I guess the first one indicates the
kind of output mode you want, and the second one indicates the raw
decoder to use? But why am I specifying 'F' if the input data is all
integers? I tried changing them respectively to 'I' and 'I;16BS', but
that didn't change the resulting blank image.

Any tips appreciated!

Thanks,

Steve

[1] http://visibleearth.nasa.gov/view_rec.php?id=8391
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100221/7dfc8eee/attachment.html>

From sflist at ihonk.com  Tue Feb 23 05:48:15 2010
From: sflist at ihonk.com (Steve Freitas)
Date: Mon, 22 Feb 2010 20:48:15 -0800
Subject: [Image-SIG] n00b Question
In-Reply-To: <1266822313.16307.414.camel@phat>
References: <1266822313.16307.414.camel@phat>
Message-ID: <1266900495.16307.1733.camel@phat>

On Sun, 2010-02-21 at 23:05 -0800, Steve Freitas wrote:

> I've verified that I'm pulling data that isn't just a bunch of zeros,
> and I've verified that the generated TIFF isn't just a bunch of zeros,
> but when I view it I get the checkerboard pattern which seems to
> indicate a translucent image. Or something.


Solved. Turns out that I needed to a) Change the mode to 'I', the raw
decoder mode to 'I;16BS' and save it as a PNG. I wonder if I just needed
to set alpha somehow on the TIFF, but no biggie, a PNG gave me what I
needed and I can use gimp to convert it to a TIFF if that's the format
my 3D scene manager needs.

Steve
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100222/0f920b64/attachment.html>

From joeoettinger at gmail.com  Tue Feb 23 11:57:55 2010
From: joeoettinger at gmail.com (joe oettinger)
Date: Tue, 23 Feb 2010 05:57:55 -0500
Subject: [Image-SIG] PIL and Python 3.1
Message-ID: <f93261ea1002230257o234f8a15gc6e8aa035d6ba452@mail.gmail.com>

First, I'd like to say thanks to  Fredrik Lundh and those working with him
for the huge and effective effort they've expended to establish and maintain
PIL.

Second, that PIL for python 3.1 would be a wonderful advance.

Personally, I just dabble in Python, but I've got an interesting story to
tell:

My son, who writes software for TI, recently wanted to help his 10-year-old
learn to program. I suggested Python. He downloaded the latest version of
Python and bought "Python Programming for the Absolute Beginner" (which uses
PIL and Pygames). He's a very busy engineer. The kid is bright, but has many
other interests. As you will have guessed, they've postponed their project
for a while.

I'm sure  that lack of support libraries like PIL and numpy has slowed the
adoption of Python 3.x. I'd guess that the uncertainty about which version
to use has even slowed the overall adoption of Python by new users.

Conversely, Python3.x with PIL and other support libraries would be
unbeatable - a beautiful intellectual achievement.

Thanks again to Lundh et al for all their selfless work!

Joe Oettinger
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100223/3c11f216/attachment.html>

From jared0x90 at gmail.com  Tue Feb 23 18:11:37 2010
From: jared0x90 at gmail.com (Jared De Blander)
Date: Tue, 23 Feb 2010 12:11:37 -0500
Subject: [Image-SIG] PIL and Python 3.1
In-Reply-To: <f93261ea1002230257o234f8a15gc6e8aa035d6ba452@mail.gmail.com>
References: <f93261ea1002230257o234f8a15gc6e8aa035d6ba452@mail.gmail.com>
Message-ID: <2960ed0a1002230911s326d2444xc40ac82aeb867d3f@mail.gmail.com>

This sort of describes the situation I have just recently ran into. I wanted
to fiddle with a wireless IP cam I picked up to monitor my driveway as the
included software was garbage and decided to try Python as I had been
wanting to learn it for sometime. Well 100 lines into my first real attempt
at a python3 program I realized that I needed PIL and had to revert back to
python2.  I know asking for something like this to be converted is a tall
tall order but as it seems to be in the works I just wanted to give it my +1
again


On Tue, Feb 23, 2010 at 5:57 AM, joe oettinger <joeoettinger at gmail.com>wrote:

> First, I'd like to say thanks to  Fredrik Lundh and those working with him
> for the huge and effective effort they've expended to establish and maintain
> PIL.
>
> Second, that PIL for python 3.1 would be a wonderful advance.
>
> Personally, I just dabble in Python, but I've got an interesting story to
> tell:
>
> My son, who writes software for TI, recently wanted to help his 10-year-old
> learn to program. I suggested Python. He downloaded the latest version of
> Python and bought "Python Programming for the Absolute Beginner" (which uses
> PIL and Pygames). He's a very busy engineer. The kid is bright, but has many
> other interests. As you will have guessed, they've postponed their project
> for a while.
>
> I'm sure  that lack of support libraries like PIL and numpy has slowed the
> adoption of Python 3.x. I'd guess that the uncertainty about which version
> to use has even slowed the overall adoption of Python by new users.
>
> Conversely, Python3.x with PIL and other support libraries would be
> unbeatable - a beautiful intellectual achievement.
>
> Thanks again to Lundh et al for all their selfless work!
>
> Joe Oettinger
>
> _______________________________________________
> Image-SIG maillist  -  Image-SIG at python.org
> http://mail.python.org/mailman/listinfo/image-sig
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100223/65293578/attachment.html>

From william.t.bridgman at nasa.gov  Fri Feb 26 14:42:11 2010
From: william.t.bridgman at nasa.gov (Bridgman, William T.)
Date: Fri, 26 Feb 2010 08:42:11 -0500
Subject: [Image-SIG] Bug Report: PIL font management clipping tops of
	characters (addendum)
Message-ID: <4057D068-DBA4-4379-85A1-EA35BEE198A9@nasa.gov>

This happens with at least
Python 2.4 (LInux) PIL 1.1.6
Python 2.6 (Mac OS X 10.5.+), PIL 1.1.6

Release notes for 1.1.7 suggest something like this might have been  
fixed in 1.1.6b1 but no clear indication in 1.1.7.

Tom
--
Dr. William T."Tom" Bridgman               Scientific Visualization  
Studio
Global Science & Technology, Inc.          NASA/Goddard Space Flight  
Center
Email: William.T.Bridgman at nasa.gov         Code 610.3
Phone: 301-286-1346                        Greenbelt, MD 20771
FAX:   301-286-1634                        http://svs.gsfc.nasa.gov/





From william.t.bridgman at nasa.gov  Fri Feb 26 14:29:43 2010
From: william.t.bridgman at nasa.gov (Bridgman, William T.)
Date: Fri, 26 Feb 2010 08:29:43 -0500
Subject: [Image-SIG] Bug report: PIL font management clipping tops of
	characters
Message-ID: <A093DC2D-66AB-4FB3-8121-B88493938BF6@nasa.gov>

Such as the circle atop the characters: ? ?, used in Angstrom.

Sample code & graphic output below.

#!/usr/bin/env python
# -*- coding: utf-8 -*-
from PIL import Image, ImageFont, ImageDraw

titleFont=ImageFont.truetype('HelveticaBold.ttf',32,encoding='unic')
size=(128,128)
image=Image.new('RGB',size,(0,0,0))
draw=ImageDraw.Draw(image)
teststr=unichr(0x00e5)+unichr(0x00c5)
textsize=draw.textsize(teststr,font=titleFont)
xposition=0
yposition=textsize[1]
draw
.text((xposition,yposition),teststr,font=titleFont,fill=(255,255,255))

image.save('fonttest.tif','TIFF')


-------------- next part --------------
A non-text attachment was scrubbed...
Name: fonttest.tif
Type: image/tiff
Size: 49280 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/image-sig/attachments/20100226/49f18ae1/attachment-0001.tif>
-------------- next part --------------



Thanks,
Tom
--
Dr. William T."Tom" Bridgman               Scientific Visualization  
Studio
Global Science & Technology, Inc.          NASA/Goddard Space Flight  
Center
Email: William.T.Bridgman at nasa.gov         Code 610.3
Phone: 301-286-1346                        Greenbelt, MD 20771
FAX:   301-286-1634                        http://svs.gsfc.nasa.gov/





From William.T.Bridgman at nasa.gov  Fri Feb 26 15:06:52 2010
From: William.T.Bridgman at nasa.gov (Bridgman, William T.)
Date: Fri, 26 Feb 2010 09:06:52 -0500
Subject: [Image-SIG] Bug Report: PIL font management clipping tops of
	characters (addendum)
Message-ID: <7F24AB15-5A47-47A8-B777-B5D1E20EE93B@nasa.gov>

Also happens with
Python 2.5.1 (Mac OS X 10.5.+), PIL 1.1.7

This happens with at least
Python 2.4 (LInux) PIL 1.1.6
Python 2.6 (Mac OS X 10.5.+), PIL 1.1.6

Release notes for 1.1.7 suggest something like this might have been  
fixed in 1.1.6b1 but no clear indication in 1.1.7.

Tom
--
Dr. William T."Tom" Bridgman               Scientific Visualization  
Studio
Global Science & Technology, Inc.          NASA/Goddard Space Flight  
Center
Email: William.T.Bridgman at nasa.gov         Code 610.3
Phone: 301-286-1346                        Greenbelt, MD 20771
FAX:   301-286-1634                        http://svs.gsfc.nasa.gov/