[ ] Give an option to add BG to images w/o BG:
- White BG
- Black BG
[x] Remove Images which do not meet a minimum size requirement
[x] Add a CLI/.bat
[x] Add a method to change that cropping from square to any shape
[ ] Add a .txt that stores the data that the person puts
[x] Added a Logging element and a relative txt file in sdh_log.txt
[ ] Add a check inside the logging txt that shows all the inputs done by the user
[ ] Do the same for "Check Choices" also
[x] Add a gradio GUI
[x] Add a GUI Based cropping tool for manual cropping of images
[x] hard bound to the dimensions given
[ ] Add a image parameter check tool inside gradio (Later)
[x] Add a loading bar
[x] Maintain resize code inside of the cropping tool
[x] Make it so that the images from inside folder "folder_loc" move onto "upload images from folder" and are applied to the image box via "load images"
[x] Fix the incorrect cropping shown in the image box
[x] Also, fix the size of the input and output box so that they don't flare up as such (height, width)
[ ] Find out why 240/236 were causing an issue, the issue, as i see it, is in detect_anime, inside the face detector, as far as i remember, similar issue on realistic images also
[x] Test out the optimised code
[x] Make the code more flexible by making it run the try-except block 1-2 times, if it still fails, then it should just move on or move the image to a different folder
[x] Add a final print the log telling how long it took
[x] Add the resizing tool inside the manual cropping tab
[x] Fix the rectangle cropping issue sigh
[ ] turn this into a pip package
[x] Make readme for gradio [x] Shift readme for ipynb to a different folder
[x] Disable Options for Suitability Check
[ ] Make a fix for the CMD closing on crash
[x] Add a dropdown for image crop type, make a place to enter custom, ukw, make a height width slider, solves dimensions and crop type
- 276, 284, 177 [x] Add size calculator for ratio [x] Add a resize to option next to this [x] Add it to the functions [x] Turn squareCrop and rectangleCrop into one function using a resizer variable for height and width instead of of making different functions
[ ] Add a image captioning service
[ ] Add VectorDB for easier retrteival of hash's of images stored
[ ] Add a try/except for files that were not able to be processed, o/p to CLI