An improved visual inspection system using visual servo

In this paper we present an improved automatic visual inspection system. In this system, homography based visual servo is used to accurately locate the camera position and attitude so that a template matching inspection can be realized. To improve the visual servo system's performance, we propose a combination strategy of a GPU based Efficient Second-order Minimization (GPU-ESM) algorithm and a GPU based Scale Invariant Feature Transform (SIFT) algorithm. By utilizing the parallel processing capability of a GPU, system processing speed and reliability has been greatly improved. In this paper, the implementation details of the visual inspection system are presented. The performance of the proposed combination strategy is evaluated with experimental data.