Automatic Inspection and Processing of Accessory Based on Vision Stitching and Spectral Illumination

Abstract

The study investigates automatic inspection and processing of the stem accessories based on vision stitching and spectral illumination. The vision stitching mainly involves algorithms of white balance, scale-invariant feature transforms (SIFT) and roundness for whole image of automatic accessory inspection. The illumination intensities, angles, and spectral analyses of light sources are analyzed for image optimal inspections. The unrealistic color casts of feature inspection is removed using a white balance algorithm for global automatic adjustment. The SIFT is used to extract and detect the image features for big image stitching. The Hough transform is used to detect the parameters of a circle for roundness of the bicycle accessories. The feature inspections of a stem contain geometry size, roundness, and image stitching. Results showed that maximum errors of 0°, 10°, 30°, and 50° degree for the spectral illumination of white light LED arrays with differential shift displacements are 4.4, 4.2, 6.8, and 3.5 %, respectively. The deviation error of image stitching for the stem accessory in x and y coordinates are 2 pixels. The SIFT and RANSAC enable to transform the stem image into local feature coordinates.

Keywords

• Illumination
• Automatic Inspection
• Light Source
• Vision Stitching