Abstract: Abstract: Drying is an important method for agricultural products processing. It can reduce the moisture content of agricultural products to a certain extent and extend the shelf life. But irregular deformation resulting from drying process will cause inconvenience for subsequent processing. In order to study the regularity of deformation during the drying process of potato slices, we built an image acquisition platform based on the Kinect sensor. Firstly, we verified the accuracy of the depth detection of the Kinect image acquisition platform by cubes with 5, 8, and 10 mm sides. The results showed that the accuracy can reach 2 mm. Secondly, we selected potato as the research object and dried them using a tunnel hot air dryer. Controlled potato slice thickness was 1 mm, drying room humidity was 15%, hot wind speed was 3 m/s to study the deformation regularity of potato slice at different temperatures (50, 60, 70, 80 ℃). After the drying process began, the potato slices were taken out of the drying chamber and put on the Kinect image acquisition platform to acquire depth and color images, then weighed every 10 minutes. We used the Kinect SDK function to achieve a one-to-one correspondence between color images and depth images. According to the position of the material in the color image, the region where the potato slices were located of interest was established, and the potato slices were located at the same coordinate position in the depth image. Gray value stretching, threshold segmentation, and edge denoising were performed on the corresponding region of depth images. Then feature extraction was used to distinguish every potato slice and calculate its shrinkage rate, mean depth values and standard deviation. The mean depth value can reflect the curling of the potato slices during the drying process. The shrinkage rate could reflect the shrinkage characteristics of the potato slices in the drying process. And the standard deviation of the depth value could reflect the surface flatness of potato slices in drying process. Then we drew the curve of dry basis moisture content and the three parameters under different temperature conditions, which could be more directly to observe the effect of temperature on the deformation of potato slices during drying. The results showed that temperature had a significant effect on the shrinkage, surface curl, and surface flatness of the potato slices in the drying process (P<0.05). With the increase of temperature, the shrinkage of potato slices increased gradually. At 50 ℃, the shrinkage rate was 54.97%. When the temperature rose to 80 ℃, the shrinkage rate increased to 64.55%. With the increase of temperature, the variation of curl and flatness of the potato chips was small. The mean depth value of the potato slices was 27.81 mm at 60 ℃, and it decreased to 18.86 mm at 80 ℃. At 50 ℃, the standard deviation of potato slices depth was 7.99 mm, which decreased to 5.71 mm at 80 ℃. Finally, using MATLAB software to display three-dimensional graphics of potato slices in five different time periods, the surface deformation of potato slices could be clearly observed. It was illustrated that the Kinect image acquisition platform could be applied to the study of deformation regularity in the drying process of potato slices, and provide technical basis for intelligent control of the drying process.