Research On Workpiece Image Recognition Technology Based On Cloud Model

Six years have passed since the concept of "industry 4.0" was proposed,The world caused the upsurge of the intelligent times of the fourth industrial revolution.Industrial robot is an important field of intelligent production.It makes industrial robot more intelligent in automatic production system by means of machine vision.The main content of this paper is to add a pair of "eyes" for the machine to identify the workpiece--research on the workpiece image recognition technology based on the cloud model.Firstly,the definition of image segmentation and traditional segmentation methods are introduced,and the algorithm implementation of image threshold segmentation is introduced.Because of its simple calculation and easy implementation,this paper chooses to improve the image segmentation algorithm based on threshold segmentation.Considering the inevitable ambiguity and randomness in the traditional image segmentation process,the cloud model combines the traditional fuzzy set theory with the probability and statistical knowledge to express the ambiguity and randomness in the uncertainty.Based on this feature,the cloud threshold-based image threshold segmentation algorithm proposed in this paper can effectively improve the accuracy of image segmentation and paving the way for subsequent image recognition.Secondly,this paper uses the geometric features of the graph,the SUSAN corner detection algorithm,and the improved Hough algorithm to extract the feature parameters of the graph,and finally realizes the recognition and classification of the geometric artifacts based on the extracted feature parameters.And the algorithm can be applied to the target recognition and positioning in the vision system of the industrial robot for the determination of the coordinates and the rotation angle of the center point of each workpiece.The positioning is captured according to its posture,and the accuracy of the recognition is for machine vision.Application is of great significance.Finally,through the comparison of experimental results,it can be concluded that the cloud image-based workpiece image recognition method proposed in this paper can effectively identify geometrical artifacts and has good stability.This method solves the neglect of factors such as fuzziness and randomness in traditional image processing,and provides a new idea for geometric artifact image recognition algorithm.