Monocular Visual Of Three-dimensional Measurements In Robotic Sorting System Applications

With the continuous progress of science and technology, the continuous improvement of the level of production line automation, robot sorting technology has also been widely used.Traditional robot sorting technology is mostly completed by the way of teaching, once the workpiece position changes will lead to the failure of the robot grab. Industrial robot sorting technology and visual technology together, so that the robot has a similar human eye function,not only to make the robot to adjust the external environment changes, but also improve the flexibility and adaptability of the robot.According to the sorting of the clutter in the industrial production line, this paper presents a sorting method combining machine vision with industrial robots. At first, this paper analyzes the imaging model of the camera, and the Zhang Zhengyou calibration method to calibrate the camera and robot hand eye relation was established, solving the camera intrinsic and external parameters and transform matrix. Then through the industrial camera to collect the image of the workpiece, the collected image smoothing, denoising and other preprocessing, in order to eliminate the noise in the image, and then use the improved Canny operator to extract image edge points and edge points x, Y direction of the gradient value as the matching information.The gradient vector normalization method is used as the similarity criterion to judge the similarity between the template image and the search image. The template matching is done to get the 2D coordinates of the grab points. Finally, the depth information of the measured object is obtained by the method of laser and vision, and the 3D position of the grasping point of the workpiece is obtained.The experimental results show that this matching algorithm can be successfully used in the clutter of the workpiece to match the target workpiece, and then through the laser and visual way to get the matching workpiece grab points of the three-dimensional parameters, and the parameters of the information transmitted to the robot controller, so as to guide the robot to accurately grasp.