• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Photoshop - Outputting Guide Coordinates

C

CycloWizard

Lifer
I'm trying to process a very large number of images (~8,000 or so). I have implemented actions in PS to do most of the work, but there's one very time-consuming part left: I have to go in to each image, map out bounding boxes (which I do using guides), then manually copy down the locations of the guides into an input file for processing the images the rest of the way using MATLAB. My question is: is there is a way to have PS automatically output the guide locations to a file? If there is, this will probably save me about 3 weeks' work and I'll buy someone a beer, but as yet I haven't been able to find a way. Even if it's a program I have to pay for (cheap program, anyway), then that's cool. I'm using PS 7 and also have access to PS 9 (CS2). :beer:
 
Because there is no good way to automate the process, since it is different in each image. Here's what the MATLAB does currently:

1. Converts the RGB input image to grayscale.
2. Rotates it to the proper orientation.
3. Finds the edges using a Canny filter.
4. Keeps only those edges within the specified bounding boxes (of which there are two per image, since there are two surfaces that I'm interested in).
5. Fits the surfaces' edges with various functions.

The bounding boxes are specified to exclude things in the image that are not actually parts of these surfaces. I've tried excluding them using skeletonization and other automated object detectors, but haven't had much success, so I've just formed the bounding boxes by hand in photoshop. It takes me about 3 seconds to put all the guides in place, but then I have to manually read the pixel locations of each of the guides and record them. This takes a long time, since there are 9 guides per image times about 8,000 images. By cropping the images properly, I have reduced the number of guides to 5-7 per image, depending on where the interfering objects appear in the raw image, but it's still ridiculous that I can't output these values automatically somehow.
 
Can you post an example image? I bet i have some code that will do what youre doing in PS. Since coming to mit i have written a number of segmentation algorithms. What distinguishes your exlcusion zone from the surround?
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top