Research On Target Recognition And Location In Industrial Robot Based On Binocular Stereo Vision

Machine vision technology has became a significant part of the modern intelligent logistics,which greatly improves the ability of real-time and accuracy of external information of logistics system,and effectively guarantees the efficiency of robot.Nowadays,automatically sorting,stacking and carrying cigarettes by robots became the main development trend of tobacco logistics.Machine vision technology is the core of this key link,which has a decisive impact on the efficiency and accuracy of automatic operation.On the basis of binocular stereo vision and the automatic control technology of industrial robot,this paper studies the automatic palletizing system that taking the strip tobacco as target object,which has certain practical application value to the automatic palletizing system of tobacco logistics.Firstly,the whole scheme of target recognition and location based on industrial robot was designed.Then according to this scheme,the forward kinematics and inverse kinematics of 6R industrial robot were studied,and the binocular camera was calibrated to obtain its internal and external parameters.In view of the problem of noise interference in the original image collected by the camera,this paper designs the image preprocessing process of average graying,piecewise linear enhancement and median filtering by comparing different methods in the image preprocessing stage.Then,in the research of feature point matching,aiming at the shortcomings of scale-invariant feature transform,speeded up robust features and other feature extraction algorithms,such as long matching time and low matching accuracy,this paper proposed a surf matching algorithm based on adaptive threshold.In this method,the basic matrix F of polar constraint relationship was obtained by using low threshold and polar constraint conditions.Then,combining with the adaptive threshold method,a lot of high-precision matching point pairs were screened out.Finally,the false matching point pairs were further eliminated through the distance and angle consistency discrimination method of feature point pairs,so as to improve the matching accuracy and efficiency.The experimental results showed that the average matching accuracy of the algorithm is more than 90%.Finally,according to the industrial robot target recognition and positioning system based on binocular stereo vision,the whole recognition and positioning system wasproposed,and the presented algorithm and the designed system were verified.The experimental results show that the target recognition and positioning system of industrial robot based on binocular stereo vision designed in this paper has the advantages of high efficiency and accuracy,and has a certain practical application value for the automatic stacking system.Finally,according to the design of the industrial robot target recognition and positioning system based on binocular stereo vision,the whole recognition and positioning system is completed,and the algorithm proposed in this paper and the designed system are verified.The experimental results show that the target recognition and positioning system of industrial robot based on binocular stereo vision designed in this paper has the advantages of high efficiency and accuracy,and has a certain practical application value for the automatic stacking system.