Chromatic Extraction & Object Detection Intern for Data Labeling
Worked on developing a chromatic extraction predicate model for delivery challan images using advanced image processing techniques. Labeled and extracted relevant objects and features in images with varying HSV layers to optimize a data retrieval system for business operations. Built a feature-based object detection application using OpenCV, SIFT, FLANN, and RANSAC, handling large volumes of daily images for accurate mapping and data extraction. • Developed efficient methods for labeling and extracting objects from images with various color and pressure layers. • Integrated object mapping and overwriting stamps to vendors with high accuracy and fast processing times. • Designed and maintained an extraction pipeline handling over 10,000 images daily, achieving high extraction efficiency. • Utilized Grid-based-median filtering to analyze and annotate image patches for enhanced object detection results.