This document summarizes research on using image processing and machine vision techniques to recognize weeds for robotic weed control applications. The researcher used a Microsoft Kinect V2 sensor to capture color and depth images of plants, then applied preprocessing, segmentation, clustering, and feature extraction algorithms to the image data. Classification algorithms were used to identify plants as weeds or crops based on extracted features like size, height, leaf shape, and position. The results showed over 90% accuracy in segmenting, localizing, and identifying lettuce and broccoli plants of different sizes. Future work will optimize and improve the algorithms to handle more plant varieties and field conditions.