As a photographer, I like to provide my readers with some EXIF-Information about my photos. Besides the informative value of EXIF photo captions, it also adds whitespace between the photos which I find much more pleasant to look at than the default way steemit.com displays photo galleries. However, especially when writing longer posts, nobody would want to add EXIF Captions for 20+ photos manually.
1/640s; 140mm; f/5.6; ISO 640
I didn't have a photo of a python, so here are some monkeys that I captured at Mt. Popa in Myanmar three days ago. Note the short exposure time that I needed because they were moving so fast and that you can now see in the cool EXIF caption - awesome, isn't it?
Repository
https://github.com/holgern/beem
https://github.com/python-pillow/Pillow
What Will I Learn?
- How to read and format EXIF-Data with Pillow
- How to upload photos to steemitimages using
beem.imageuploader - How to create cool EXIF captions
- How to add a watermark with Pillow
Requirements
- Working Python installation
- Beem Python library by
with a valid posting key added to the CLI wallet and the environment variable
UNLOCKset to the CLI wallet passphrase PillowPython libraryfractionslibrary (comes with Python) for EXIF calculations
Difficulty
- Intermediate
Tutorial
In this tutorial you will learn how to write a Python script that will read the EXIF-Data of all images in a folder, upload them to steemitimages and output a convenient HTML-snippet for each photo including the Exif caption.
Before
After
Reading EXIF with Pillow
Before we start, we need to define a default function for Pillow for reading the EXIF of a photo:
def get_exif (exif,field) :
for (k,v) in exif.items():
if TAGS.get(k) == field:
return v
Now, we open the image and call our get_exif() functions with the parameters we want to know:
exif = img._getexif()
tmp_exp = get_exif(exif,'ExposureTime')
tmp_focal = get_exif(exif,'FocalLength')
tmp_fstop = get_exif(exif,'FNumber')
ISO = "ISO "+str(get_exif(exif,'ISOSpeedRatings'))
Formatting the EXIF outputs
You may have wondered why we called the variables tmp_. Let's quickly print out the results with print(tmp_exp, tmp_focal, tmp_fstop, ISO) to see what is wrong with them:
(1, 50) (280, 10) (18, 10) ISO 560
This is not what we want to get! The numbers are correct though, we just have to format them correctly. To prevent empty values from crashing the script, we use proper exception handling that returns the default value "None" if the EXIF field in question is empty or could not be processed:
try:
ExposureTime = str(Fraction(tmp_exp[0]/tmp_exp[1]).limit_denominator())+"s"
except:
ExposureTime = "None"
try:
FocalLength = str(int(tmp_focal[0]/tmp_focal[1]))+"mm"
except:
FocalLength = "None"
try:
FNumber = "f/"+str(tmp_fstop[0]/tmp_fstop[1])
except:
FNumber = "None"
Let's now create a string out of all the values. If the value of any of the EXIF data is "None", we return an empty string.
settings = "" if ExposureTime == "None" or FocalLength == "None" or FNumber == "None" or ISO == "None" else ExposureTime+"; "+FocalLength+"; "+FNumber+"; "+ISO
With our example values from above, we now get the following output:
1/50s; 28mm; f/1.8; ISO 560
That looks much better!
Uploading photos with beem.imageuploader
Uploading the photos to steemitimages is very easy thanks to beem:
steem = Steem()
iu = ImageUploader(steem_instance=steem)
steem.wallet.unlock(walletpw)
postimage = iu.upload(path+"/"+imgurl, postaccount)
Now, we only need one line to print out a HTML snippet for each image containing the image that we just uploaded to steemitimages and its EXIF data that we previously saved into the string settings as a caption.
print("<img src='"</span><span class="ll-pct">+</span><span class="ll-nam">postimage</span><span class="ll-pct">[</span><span class="ll-str">'url' ]+".jpg' />
"+settings+"")
Finishing the function get_image()
Now, let's wrap this code in the function process_image():
def process_image(imgurl):
try:
"""
Reading the Exif-Data with Pillow
"""
img = Image.open(path+"/"+imgurl)
exif = img._getexif()
tmp_exp = get_exif(exif,'ExposureTime')
tmp_focal = get_exif(exif,'FocalLength')
tmp_fstop = get_exif(exif,'FNumber')
ISO = "ISO "+str(get_exif(exif,'ISOSpeedRatings'))
"""
Formatting the EXIF-Data
"""
try:
ExposureTime = str(Fraction(tmp_exp[0]/tmp_exp[1]).limit_denominator())+"s"
except:
ExposureTime = "None"
try:
FocalLength = str(int(tmp_focal[0]/tmp_focal[1]))+"mm"
except:
FocalLength = "None"
try:
FNumber = "f/"+str(tmp_fstop[0]/tmp_fstop[1])
except:
FNumber = "None"
settings = "" if ExposureTime == "None" or FocalLength == "None" or FNumber == "None" or ISO == "None" else ExposureTime+"; "+FocalLength+"; "+FNumber+"; "+ISO
"""
Uploading to steemitimages using beem
"""
steem = Steem()
iu = ImageUploader(steem_instance=steem)
steem.wallet.unlock(walletpw)
postimage = iu.upload(path+"/"+imgurl, postaccount)
"""
Printing out the result
"""
print("<img src='"</span><span class="ll-pct">+</span><span class="ll-nam">postimage</span><span class="ll-pct">[</span><span class="ll-str">'url' ]+".jpg' />
"+settings+"")
except Exception as error:
print("Error during image processing "+repr(error))
Queuing Images
Finally, let's write the main function. It calls our function process_image() for each file in our folder path with the filename as parameter. We could also add a filter to only queue files with image filetypes, but this would go too far for this tutorial.
if __name__ == '__main__':
for image in os.listdir(path):
process_image(image)
print("Finished")
Remember to also set your own values on the top of the script for path (path to your folder that contains the image files) and postaccount (your steem account previously added to the beem wallet). Remember to set the environment variable UNLOCK or alternatively define walletpw as your beem CLI wallet password.
path = '00steemit'
postaccount = 'yoursteemaccount'
walletpw = os.environ.get('UNLOCK')
Now we can run the script and get an output similar to this:
<center><img src='https://cdn.steemitimages.com/DQmXmWn4NaBzUaKhEGru2hgVrcMXbCE86r49SLUuqeeWp4P/image.jpg' />
1/50s; 28mm; f/1.8; ISO 560</sup></center>
<center><img src='https://cdn.steemitimages.com/DQmYNcx8Tzx8tVHEqubqPNrGY6movmdzDVdEPq53ALZ56pR/image.jpg' />
1/30s; 20mm; f/1.8; ISO 5000</sup></center>
<center><img src='https://cdn.steemitimages.com/DQmP6LsHirykSpubdzW2NXu13LSZQecDQ41fpbuhxA7fpma/image.jpg' />
1/500s; 19mm; f/1.8; ISO 900</sup></center>
Finished
Bonus: More fun with Pillow
Pillow can do much more than just reading EXIF-Data, it is also capable of editing images.
I use it to add a watermark to my photos: Instead of simply uploading the photo with postimage = iu.upload(path+"/"+imgurl, postaccount), we make a watermarked copy of the photo in a second folder:
watermarkedpath = '00steemit'
watermarkpath = 'watermark.png'
watermark = Image.open(watermarkpath)
destination = img.copy()
position = ((destination.width - watermark.width - 32), (destination.height - watermark.height - 30))
destination.paste(watermark, position, watermark)
waterimgurl = watermarkedpath+'/'+imgurl
destination.save(waterimgurl)
postimage = iu.upload(waterimgurl, postaccount)
Of course, you could also read more EXIF Data such as the lens used with Pillow, add the title as an ALT-Attribute to the image for SEO optimation or link each image to its steemitimages URL for a fullscreen preview.
I run the script from a folder containing the script itself, the folder 00steemit where I put my photo files, the optional watermark file watermark.png and the optional folder for watermarked images 01steemit. I run the script with python exifcaptions.py > output.txt which saves the HTML snippets generated into the file output.txt which contents I then copy into my steemit post.
Enjoy the cool EXIF captions!
Proof of Work Done
You can find the complete code on Github:
https://github.com/tiotdev/steem-tutorials/blob/master/cool-exif-captions/exifcaptions.py
I would love to see this functionality implemented in some Steem frontends such as 
