Research On The Visual Identification And Positioning Of Industrial Robots' Operation Targets

Along with rapid development of science and technology.The automation of industrial robots which has made a further development.However,in a complex factory environment and the interference of factors such as the size and shape of the operation target,the level of intelligent flexibility of industrial robots is low.How to achieve the goal of rapid robot identification and positioning of industrial robots has become an urgent problem to be solved in the intelligentization of industrial robots.For this reason,this paper combines robotics and machine vision technology,and studies the identification and positioning of intelligent flexible operation targets for industrial robots in complex environments,which is of great significance for further improving the intelligence of industrial robots.The main research work of this article is as follows:1)According to the influence of the image quality degradation caused by the uneven surface of the metal member,this paper proposes a histogram correction method based on color space.Through the Y?CbCr conversion of the color space,the histogram equalization of Y processing is performed on the components,which effectively reduces the brightness of the high-luminance area and maintains the brightness of the low-luminance area.At the same time,it uses filtering and sharpening to further improve the image quality.2)Aiming at the fast,high-precision identification and matching of job targets,through the research of algorithms and algorithms,this paper proposes a method of matching SURF algorithm,BRIEF algorithm and Hamming distance.The feature points of the image are detected by the SURF algorithm,and the descriptors are used to describe the feature points.The initial matching between the template and the job target image is achieved.Then the Hamming distance is used as the matching similarity measure,and the secondary precision matching is performed using the RANSAC algorithm.Improve the accuracy of the match.3)According to the relationship between template and job target image,a 4-parameter affine transformation model is established.The sub-pixel edge features of Zernike moments are extracted from the template image based on Canny operator.The morphological filling algorithm is used to obtain the center-of-mass coordinates of the template image,and the affine transformation is used to solve the center-of-mass coordinates of the target.Simultaneously with the minimum external moment,the two-dimensional dimension of the job target is measured.4)Based on the analysis of system configuration and hardware performance parameters,a machine vision-based robot operation target recognition experimental platform was constructed.Visual system calibration was completed without considering camera distortion.Through the use of explorer robot structural components as the target of work,the experiments of target identification,positioning and two-dimensional measurement were completed,and the feasibility of the proposed algorithm was verified.