This paper introduced the texture feature analysis method in digital image processing technology into laser biological tissue welding, and the microstructural changes of laser-welded skin in vitro were deeply analyzed. The second-order statistical features calculated using gray co-occurrence matrix were used as primary quantitative indexes to characterize the microstructural characteristics of isolated skin. The effects of laser parameters and their interaction on the microstructural and texture characteristics of laser-welded skin in vitro were obtained using a response surface method. The results show that the laser scanning speed is the key factor that affects the microstructure of isolated skin subsequent to welding, and the angular second moment, correlation, entropy, and contrast have significant effects on the texture feature parameters. Furthermore, the laser power and the interaction between laser power and welding speed have significant effects on the angular second moment, correlation, and entropy.