Method of ripe tomato detecting for a harvesting robot

Abstract

Harvesting robots could be efficiently implemented in a greenhouse for picking ripe tomatoes. For such robots, tomato detection is a basic operation of visual control which provides the correct target object for precise moving of the end-effector or gripper. The tomatoes were chosen to perform this research work not only because they occupy the second place in agriculture after potatoes, but also due to their simple round shape and the apparent red color of the ripe fruit. In this study, an algorithm for automatic detection of ripe tomato and discrimination it from the natural background is proposed. The proposed algorithm is based on image processing procedures and starts with translating original color photograph to grayscale image for further treatment. An algorithm of red tomatoes detection was developed and verified using special functions and inbuilt scripting language of Matlab software.