An automatic method for identifying the microscopic feature of ray-cell number

In order to improve the efficiency and objectivity of identification the characteristics of average cell number of coniferous wood and reduce the subjectivity of manual identification of the characteristics, an automatic identification method for the average number of ray cells in coniferous wood is proposed in this paper. Based on this automatic identification method, the microscopic image of coniferous wood is grayed, binarized and denoised successively. The cell number is obtained by labeling the connected region of the image. The Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm is improved for counting the ray number. Finally, the average number of ray cells and its characteristics are obtained. Experimental results show that the accuracy of this method is 65%, and the method is feasible, but it still needs to improve the effect of binary image denoising and clustering accuracy, and improve accuracy rate of this method.