In this paper, semiconductor lasers with wavelength of 650 nm，powers of 13 and 25 mW, respectively, were used as light source to illuminate the surface of peaches (early harvested one, Xiahui and middle-late harvested one, Zhaohui) during their shelf life. The images gained by the computer vision were pre-treated to get area S1 and S2 and then the captured images were analyzed. The relationships between the image parameters and the correlation coefficients between qualities parameters of peaches were investigated. We found that the parameter AS1-AS2 had a better correlation with the maximum puncturing force, CIE a*, hue angle H° and SSC/TA. The results proved the possibility of non-destructive detection of peach quality by using laser instrument.