| In the context of the development of "Made in China 2025",intelligent assembly technology is one of the most popular research projects that uses information technology to promote industrial change.But intelligent assembly technology still faces many challenges.Metal materials in assembly parts are prone to data loss due to high light areas on the surface,which in turn affects the accuracy of three-dimensional matching,making it more difficult and less accurate to obtain three-dimensional information of the workpiece.The following research has been conducted on binocular vision related technologies for assembly robots.(1)For the problem of highlights on metal surfaces,a restoration algorithm based on chromaticity space conversion and BSCB is proposed.The image is converted in RGB and CIE-XYZ chromaticity space,and the highlight area is derived as the input of BSCB algorithm mask for restoration.The algorithm effectively removes the saturated pixels of the image and the highlights are removed more thoroughly.(2)The Sum of Absolute Differences(SAD)algorithm in the local stereo matching algorithm is studied and analyzed,and the edge detection is integrated into the SAD algorithm and fused with the Census generation value weighting.Finally,an experimental comparison with the traditional SAD algorithm and Census algorithm is conducted.The running time of this experimental algorithm is 6.80 s,and the false matching rate is reduced by 35.7% and 36.6% respectively compared with the other two algorithms,indicating that the algorithm is better than the other two in terms of matching accuracy and can well ensure the real-time performance.3)Build the assembly robot with binocular vision system.The 3D information of the workpieces was obtained through experimental calculations and compared with the actual3 D information,and the error range was controlled within 0.5mm,meeting the requirements of factory assembly,and the effectiveness of the algorithm was verified.Figure 42;Table 12;Reference 47... |
PDF Full Text Request