Introduction

The agriculture field is working to save our environment by making certain exceptional changes in its practices to secure crop production. This approach is totally based on technology which can support soil preparation, planting and weed eliminating process. The most vital problems surface due to weeds which increase the biological competition with existing crop such as excess use of fertilizer, water and manual working hours. Manual work of weed inspection and removal takes a longer time. However, robotic applications are able to perform weed detection in excellent way. It is one kind of mechanical elimination task. Weed area classification using images taken in outdoor environment is very complicated due to random and uncontrolled light conditions. But images are evaluated through image processing techniques such as shape, size and texture features which make the classification of weed area easy. A strategy to employ cameras is required to determine the positions of weed in crops so that stereoscopic images can be processed. The image pixels are processed through various image processing algorithms which helps to differentiate between crops and weeds. Once the weed position accurately mapped, it is removed by robotic arms. This kind of automatic weed detection robotics work is based on sensors and image processing algorithm (Figure 3.1).