The Research And Achievement Of Log Diameter Class Identification Algorithm Based On End Image Processing

The detection of log volume has an important position in modern wood industry, at the moment, it often uses traditional manual work to measure logs, its operating efficiency is lower and calculation precision is worse, it is more difficult to guarantee the measurement accuracy of results. With the development of wood industry, the traditional method of log scaling can not meet the demand of precision. Therefore, we must study a new, high-precision digitized measurement method, and fit the development of modern wood industry.The paper is a part of "Log Diameter Class Video Identification and Import of Digital Log Scaling Technology" which is supported by the fund of "948" project of State Forestry Administration, it mainly researches the key technologies which are involved in the process of digital log scaling, solves the problem of the detection of log end face and diameter class, proposes algorithm, achieves the calculation of log volume preferably.In order to identify the diameter class better, first of all, it needs to process the end faces of logs. According to the feature of the image of log ends, removed internal hole noise and external band noise of the image by combining the area threshold method and mathematical morphology. Processed the image by the method of homomorphic enhancement for the first time, transformed the image to the logarithm form, and applied Fourier Transform, then connected low frequency component with irradiated component, and high frequency component with reflex component, with this approach, the interrelated contrast of image can be enhanced, and the precise contours of the log ends can be extracted.Log end is similar to ellipse, in the course of diameter class recognition, this paper presents the method of diameter class recognition, proposes an algorithm of calculating the initial values which used in ellipse fitting based on the least distance limit. The accuracy of ellipse fitting is largely affected by the initial value, the method received appropriate initial value by distance calculation for the boundary points of the target image; and then calculated the smallest Euclidean distance between the boundary points and fitting ellipse by using least squares principle, thus got the best fitting ellipse parameters of major axis and minor axis, which are long diameter and short diameter of log; then, according to the length of log, calculate the log volume by formulas, also, design the system.The experimental results show that the algorithm can identify the log end and diameter class better, also has higher reliability and practicability.