AN ALTERNATIVE TO THE MANUAL COWPEA SORTING PROCESS

Abstract Cowpea traders find the manual sorting process of cowpeas laborious and time-consuming. Based on the volume of cowpeas and the proportion of damaged cowpeas, the process can span a time interval of about 6 hours. This project integrates the power of computer vision and convolutional neural networks to develop a solution for the cowpea trader to effectively segregate good cowpeas from the damaged ones.