Authors: Zhichao Meng, Wei Zeng, Pengcheng Wang, Zenghong Ma, Xiaoqiang Du
Identifier: CSBE21422
Download file:
Published in: CSBE-SCGAB Technical Conferences » 5th CIGR and AGM Quebec City 2021 » World Congress on Computers in Agriculture and Natural Resources

Download RAW file:
Description: Existing tomato picking robots have low accuracy of target fruit recognition due to factors such as natural light intensity changes and excessive occlusion. The clustered tomato growing in the greenhouse is selected as the research object, and a method to identify the axis of each single fruit for a clustered tomato. Based on the tomato image collected on-site as the data source, the I color component in the YIQ color space is selected as the image segmentation factor first. Then the automatic threshold segmentation of the image is completed based on the iterative method, and then the morphological operation and Hough transform is used to identify the tomato fruit. Finally, based on the moment feature, the characterization and identification of the fruit axis is completed in the image, and the comparison with the marked area of actual fruit axis in the image is carried out. The comparison results show that the accuracy of tomato fruit recognition based on morphological operation and Hough transform is 93.2%, and the accuracy of fruit axis recognition is 85.5%, which meets the accuracy requirements of robotic picking.

Pages -
Date: 2021-06-11
Technical field:
Conference name: 5th CIGR International Conference and CSBE-SCGAB AGM 2021, Quebec City,QC, 11-14 May 2021.
Session name: World Congress on Computers in Agriculture and Natural Resources (WCCA 1)

Other information:
Type: Text.Abstract
Publication type: Presentation
Coverage: China
Language 1: en
Language 2:
Rights: Canadian Society for Bioengineering
Other files: