Archiving family photos can be a monumental task. Scanning tons of photographs, editing them, and keeping track of what information might be on the backs of those photographs is generally an incredibly tedious process. Especially when there are hundreds or thousands of photographs to scan!
Using the GIMP scripts and plug-ins described in our earlier posts, you can easily improve your throughput when dealing with these images. This post will discuss a few things to keep in mind to help make the process even faster.
The first step occurs when you are scanning the images. Say you have a bunch of photos that have information written on the backs. The easiest thing to do is to arrange multiple photographs on the scanner platen. You can then use the script mentioned in this post to separate these images automatically. Matching the fronts of the photos with the backs of the photos is still an issue.
- TIP: Scan the fronts of all the images, then turn over all the images that have text on the back. Be sure to place the images in the same place as when the fronts were scanned. This will keep the same number of photos in the scans of fronts and backs. This will help the script name the separated files correctly.
The backs of photos generally are much more white than the fronts. This may require you to play with the “Selection Threshold” value in the script. The lower the number is in this box, the fewer shades of white it will attempt to select.
- TIP: If the divide script noticeably selects multiple photos as a single picture, you can stop it and start over, but only when it is in the “selection” phase. We can purposely crash the plug-in by using the rectangle select tool and selecting any portion of the original GIMP image. This will confuse the plug-in and it will crash before it gets to the time consuming parts of opening each photo separately and saving them to file.
The “Append to File Name” feature in the script allows you to add an identifier to the images to separate the backs form the fonts, but to keep them in order in the folder they are saved in. This will help you keep track of which back belongs to which front, but the plug-in is not smart enough to know which picture and back-of-picture match each other. It is up to you to make your original scanned images set up such that the script will name the backs the same as the fronts. For instance the front of a given image is “IMAGE0007.tiff” the back might be called “IMAGE0007_back.tiff”. To do this, you must plan ahead and understand how the script looks for the next image to save. Feel free to delete duplicates of photos that you didn’t turn over.
The script looks for the image closest to the top of the screen first (top to bottom). If there are multiple images that are exactly the same height, it will choose the leftmost image first and work its way to the right (left to right). This includes ‘not-quite-whitespace’ that might surround the images.
- Tip: Be sure to make a good grid of white around all the images and move the photos on the GIMP page to make sure the names of the backs and fronts will match based on the rule “Top to bottom, Left to right”
Always double check the names after you run the script on the backs to make sure the backs match the correct fronts based on the filenames. You want the image, and the back of the image to have the same base name and number as mentioned above.
- TIP: Sometimes, deSkew doesn’t recognize the margins and skews the image even worse. This isn’t a problem, simply open the resulting image back in GIMP, and click the “Rotate” tool (or hit Shift+R). Adjust the image to be straight and save it.
Once you’ve scanned and separated the pictures and backs of pictures, and confirmed that all the images are names correctly (again fronts and backs named as mentioned above) open a particular image and its back in GIMP. Now simply combine these images to be side by side using the script mentioned in this post. You will need to install the script to the GIMP scripts folder.