Automatic recognition and grasping system based on vision and ultrasound technology
The visual sensor can visually reflect the external information of the object, but a single camera can only obtain a two-dimensional image of the object. Although stereoscopic vision can provide three-dimensional information, but the same shape, only objects with different depths are difficult to recognize (such as a hole object, a ladder) Matters, etc., and have certain requirements for ambient light. Because the ultrasonic sensor is insensitive to light, object materials, etc., the structure is simple, and the distance from the point to be measured can be directly obtained. Therefore, the method of combining visual and ultrasonic measurement is used to obtain the two-dimensional image information and the ultrasonic sensor. The depth information is integrated and inferred, and the workpiece to be assembled is automatically identified and spatially positioned, and the spatial position and posture of the actuator at the end of the robot are determined, so that the workpiece can be accurately grasped at a suitable location.
1 system principle and structure
The system consists of a robot, a CCD vision sensor and an ultrasonic sensor, and a corresponding signal processing unit. The CCD is mounted on the end effector of the robot to form the hand-eye vision. The receiving and transmitting probe of the ultrasonic sensor is also fixed on the end effector of the robot . The CCD acquires a two-dimensional image of the object to be recognized and grasped, and guides the ultrasonic sensor to obtain the depth information. . The system structure is shown in Figure 1.
system structure
Image processing mainly completes the accurate description of the shape of the object, including the following steps: a. Image edge extraction; b. Weekly tracking; c. Feature point extraction; d. Curve segmentation and segment matching; e. Graphic description and recognition . After the edge of the image of the object is extracted, the edge is refined by the contour tracking, the false edge points and the noise points are removed, and the edge points constituting the closed curve are subjected to Freeman coding, and the XY coordinates of each chain code direction and each point on the curve are recorded. Values ​​for further analysis of the geometric properties of the object. In this study, the search direction and order of edge points in the traditional weekly tracking algorithm are improved, and the method of eliminating redundant points in time is adopted in the search process, which reduces the data amount and operation time, and has better performance. Noise reduction and smoothing effects. When extracting image feature points, combining the polygon approximation method with the method of calculating curvature can overcome the shortcomings of the polygon approximation method which is easy to generate pseudo feature points and the calculation curvature method is too large. The object image acquired by CCD can be processed. Extract some features of the object, such as the centroid coordinate, area, curvature, edge, corner point and short axis direction of the object. According to the characteristic information, a basic description of the shape of the object can be obtained. On the basis of the image processing, the depth of the ultrasonic sensor to be measured by the visual information is measured, and the depth (height) information of the object is obtained, or the workpiece along the workpiece is to be tested. When the surface moves, the ultrasonic sensor continuously collects the distance information, scans the distance curve, and analyzes the edge or shape of the workpiece according to the distance curve [1]. The computer combines the visual information and the depth information to infer, and then performs image matching, recognition, and controls the robot to accurately grasp the object in an appropriate posture.
2.1 Extraction of workpiece image edges
Complex artifacts are often reflected in an image with more than one gray level, and only one gray threshold is used to extract meaningful edges.
Silicone Protective Sleeve,Silicone Sleeve For Glass Bottle,Silicone Water Bottle Sleeve,Bottle Silicone Sleeve
Nantong Boxin Electronic Technology Co., Ltd. , https://www.bosencontrols.com