Multi vision detection is an image processing technology which captures images of the goal from different angles with several CCD sensors, and obtains the information of the image after matching, segmentation and minus. The multi vision detection technology can obtain more information of the image than monocular vision, which reduces the complexity of follow-up treatment and improves the measurement accuracy of detection systems as well as the anti-interference ability and measurement efficiency. However, in order to obtain images with different information, it requires better image and lighting systems than monocular vision system. Based on the requirements of the experiment, the lighting system of multi vision detection is designed, and the experiment platform is set up. It is confirmed that the type of light source, the structure of lighting, the optical properties of detected objects, and the background characteristics are important factors for visual inspection technology, and the research provides the experimental basis for the promotion.