In order to obtain the image of seedless white grape, the influence of background paper color, light source structure and lighting mode on the image was fully considered on the self-made acquisition platform. Firstly, the optimal level of each factor is determined by single factor test, then the acquisition platform is optimized by orthogonal test, and finally the optimal value level of each factor is determined by subjective and objective comprehensive evaluation image method. The results showed that the best values of the machine vision imaging system of seedless white grape were: the background paper is white, the light source structure is strip light source -60°, the lighting mode is open field mode. Under the conditions, the image quality is the best.