Automated Pepper Berries Classification with Edge Detection and Template Matching

Pepper berries classification has been performed manually by human vision. Hence the procedure consumes a lot of manpower and time. This research discusses an application of computer vision to classify the peppercorn samples. We propose the use of edge detection and template matching from the real time gray scale images taken from the peppercorn samples. Color images and color detection is applied in this research as they consume large volume of CPU memory and reduce the classification speed in real time (Reece et al. 1998, Zou et al. 2007). The paper discusses the background of peppercorn feature classifications, followed by the peppercorn recognition algorithm, evaluation and results of the recognition.