AI That helps in Forensic Image Data

I devised an algorithm that moved a sliding window across an image, running edge detection within each window to segment the image and identify which regions were in focus versus blurred. This let the algorithm separate foreground from background based purely on sharpness, not manual selection. Once separated, I applied blurring and deblurring techniques differently to each region — restoring detail in unfocused areas that might otherwise hide important information. The project was my first hands-on introduction to spatial processing, image restoration, and image/video inpainting.