Edge detection for online image processing of a vision guide pick and place robot

Abstract

Robotics technology has merged into human's ev-eryday life, from industry to domestic application, including agriculture and creates digital farming. Agriculture industry currently faces new challenges, including aging farming and higher demand for food to feed the growing population. It is expected that robot application in agriculture can reduce the production cost in the long run and improve product quality, and increase productivity. A robot can be applied from seeding, maintenance, crop harvesting, and packing. A pick and place robot is an ideal form in accumulation and packing the harvested product. This type of robot a vision sensor to detect the crop to be collected or object to be picked and placed. However, the application of image processing needs to be kept adequate but straightforward to accommodate the current microcontroller technology. This paper discusses the implementation of an edge detection method for recognizing and tracking an object, and the target considered in this study is a tomato. Twenty experiments were conducted to pick and place a tomato from 3 random positions to a basket. The total average time of robot to finish its task is 05.81 s. The experimental result shows that the proposed method is effective for a 4DOF pick and place robot.