Development of a mobile robot with vision based perception for mechanical weed control

Abstract

Field of digital image processing is gaining attention worldwide because researchers use technology extensively in various applications in everyday life. Therefore, an investigation was carried out for the purpose of developing a computer algorithm that can identify weeds and chili plant thus reducing the work of man. The main objective is to use the concept of erosion and dilation is expected to overcome the common problems in the real-time image processing. In this research, technique of erosion and dilation has been developing to produce the set of feature vectors that can identify and classify decisions. The methodology involves has three main phases, development of the database, phase of the development of algorithms and the phase of system integration and prototype implementation. From the result 120 out of 150 sample data have proved that set of feature vector using the proposed technique successfully produce classification results.