Design And Experiment Of Picking Camellia Oleifera Fruit Based On Binocular Vision Twist-comb

Camellia is one of the unique high-quality oil crops in my country,which has high edible and industrial value.According to the existing characteristics of picking camellia oleifera fruit,combined with fruit picking experience at home and abroad,this paper proposes a twist-comb type oleifera fruit picking method based on binocular vision assisted technology.This method focuses on the twist-comb type of camellia oleifera fruit with low bud damage The picking mechanism model design,combined with image processing technology,realizes the recognition and segmentation of mature camellia oleifera fruit,and uses binocular stereo matching technology to complete the spatial position calibration of camellia oleifera fruit,which realizes the effective recognition of camellia oleifera fruit and the positioning in three-dimensional space.It is a twist comb type.Camellia fruit picking agencies provide necessary location information.The main research contents of this paper are as follows:(1)By studying the physical and mechanical properties of Camellia oleifera fruits and flower buds,and using the difference in longitudinal tensile force and twisting mechanical characteristics of Camellia oleifera fruits and flower buds,a three-dimensional model of twist-comb picking was established in Solidworks,and the whole Camellia fruit was determined to reduce the damage to the flower buds.For the picking plan,a twist-comb-type oil tea fruit picking device was designed.(2)Using the technical advantages of Zynq platform,an image acquisition system based on OV5640 binocular camera is built.The PL part of the platform realizes the transmission of image data acquisition and DMA module,which reduces the burden of CPU and improves the efficiency of image acquisition and transmission.The image data collected by the system provided the hardware basis for the segmentation and extraction of mature Camellia oleiformis fruit and binocular stereo matching.(3)Preprocess the collected images of Camellia oleifera,and analyze the preprocessed images of mature Camellia oleifera,immature Camellia oleifera,and Camellia oleifera leaves based on color difference,and propose a color difference-based threshold segmentation based mature Camellia oleifera fruit recognition algorithm,and The algorithm is used to segment Camellia oleifera images.The test results show that the algorithm can reach 75.5%～83.3% of the recognition rate of mature Camellia oleifera.(4)The binocular camera is calibrated according to the black and white chess camera calibration method,the internal and external parameters of the camera are obtained,and the rotation matrix and translation vector of the spatial position relationship of the binocular camera are obtained.The SURF algorithm was used to extract the feature points of Camellia oleifera fruit,and the mismatched feature points with limit constraints were used to complete the calculation of the three-dimensional coordinate of Camellia oleifera.Experiments show that the positioning error of the oil-tea camellia fruit space is between 1.789～6.801 cm,which meets the precision requirements of the twist-comb-type oil-tea camellia picking mechanism.