Research on State Recognition of Platen Based on Improved K-means Algorithm

The recognition rate of platen state is easily affected by illumination change or shooting angle. In order to overcome the influence of light variation or shooting angle, we propose a method of platen state recognition based on improved K-means algorithm. First, holomorphic filtering enhancement and perspective correction are performed on the collected platen images, and then the platen areas are divided equally according to the number of rows and columns of the platen. Secondly, the improved K-means algorithm is utilized to segment the platen image. Analysis to determine the status information of the pressure plate. In addition, we also conducted an experimental comparison to compare the algorithm with RGB threshold segmentation and traditional K-means clustering segmentation results. The experimental results demonstrate that the improved K-means algorithm significantly improves the accuracy of the segmentation of the platen image; and the recognition rate of the platen state reaches 99.6% by the algorithm of this paper.