Edge Detection of Variety of Cowpea Leaves Using OpenCV and Deep Learning

An edge detection algorithm using OpenCV and a deep neural network is compared in this paper. First, we have collected data (cowpea leaves) from ICAR-NBPGR, Pusa, New Delhi. Second, we have applied data preprocessing over the dataset with the help of the Keras library of deep learning. Then we implemented five different edge detection techniques i.e., SobelX, SobelY, Sobel Combined, Laplacian, and Canny algorithms. In comparison with the manual approach, the OpenCV implementation approach shows better performance. This paper analyzes how well these algorithms perform in terms of noise. Asa result, out of five algorithms canny is having low noise and clear edges of veins of the leaves. Thus, based on the results, the OpenCV canny performs better than other methods for edge detection.