Defect Detection in Solar Panels

Problem Statement: Given an image of solar panel taken through an IR camera, detect the nature, location, and the extent of the damages within the cells.

Robust computer vision algorithm to extract individual cells from a panel with 100% accuracy.
Deep learning models to segment, detect and classify the defects in individual cells with 95% accuracy.

Image Quality Assessment, Hough Transforms, RANSAC, Object Detection, Semantic Segmentation