Modeling apple quality changes based on laser scattering image analysis

650 nm, 25 mW semiconductor laser beam was applied as light source, and the quality of apples(c.v. Red Fuji and c.v. Gala) was monitored with a computer vision system. The laser scattering images of apples were analyzed. The maximum puncture force, CIE a^* value, hue angle, SSC/TA(soluble solid content/titrable acidity) ratio, and laser image parameters (S1, S2, S3, S4) were measured during the storage of apples. The results indicated that different sizes of the apples had no significant effects on the total number of pixels in the scattering image, while different sides(towards or against the sun) of apples made significant effects on the pixel numbers(P≤0.05). It was found that image parameter S3 had the highest Pearson correlations with other apple quality parameters. The stepwise regression analysis showed that the models had the best fitting(R²=0.99) when the parameter S3 was used in the model. This trial laid a theoretical basis for applying low power laser beam to monitor apples' ripeness and quality in a nondestructive way.