Enhanced Approach for Weeds Species Detection Using Machine Vision

Precision Agriculture is a clear standout of applying the most recent advances of intelligent systems. The motivation behind the adoption of such a systems is to reduce costs, increment treatments quality and efficiency, thus, increasing the quantity and the quality of agricultural products. In our study we used a histograms based on color indices to discriminate between three classes: soil, soybean and broad-leaf(weeds). This feature representation was tested with two classifiers Back-propagation neural network (BPNN), and Support Vector Machine (SVM). Our approach achieved a state of the art performance with an overall accuracy of 96.601% for BPNN, and 95.078% SVM.