Using Image Processing Technology for Detection of Red Grape Fruits in natural conditions

Abstract

With the increase of greenhouse cultivation of grapes in order to reduce the labor cost, the mechanization of harvesting operations is one of the important issues in recent years. In this research, an image processing algorithm was designed so that the robot on the tree could recognize the position of red grapes. In order to design the algorithms, a total of 500 product clusters were taken in natural conditions. The color information of each image component containing clusters from the red grape background (leaf, sky and ground) was extracted and analyzed. RGB color space was used to identify clusters of red grapes. The results showed that 98% of the design area of the algorithm is able to extract grape clusters in a photo.