Abstract: To solve the problem of component identification in automatic disassembly, a typical components identification method is proposed based on the depth image of Kinect and point cloud features. Firstly, a nonlinear filtering algorithm is used to process the acquired depth image and obtain the optimized point cloud target, and an eight neighborhood depth difference algorithm is proposed to extract the point cloud edge. The segmented point cloud edge is then detected using the RANSAC algorithm, and edge features are extracted to identify components. This method can identify typical components, and experimental results verify the effectiveness of the proposed method.