Research On Log Volume Measuring And Implementation On Android System

Log volume measurement is a key link in timber production and processing.With the development of science and technology,the accuracy of log volume measurement is also increasing.For a long time,the measurement of log volume is basically the method of manual measuring.This method is easily influenced by the working attitude and experience of the inspecting staff,and it has low efficiency,long period,poor accuracy and high cost,and the result is not accurate enough.In this paper,we analyze the characteristics of equal length pile logs,and puts forward a method of log volume detection based on computer vision,and make software implemented on Android mobile phone.In order to get the log image,in order to get a clearer detail of the log edge,it is necessary to shoot the log image near the short distance,but it is difficult to take the bundled logs to take the picture in the short distance shooting,so this paper uses a close view to shoot a number of pictures and then uses the image stitching technique to get the panorama of the log image.In this paper,an improved Harris algorithm is proposed for corner detection.First,the original gradient operator is changed to improve the detection sensitivity of the operator.A new corner response function is used,and no artificial parameters are needed.The image registration is performed by the normalized cross correlation matching method in the process of image stitching;A consistent sampling method is used to eliminate mismatches and achieve seamless image stitching.The traditional Canny algorithm is sensitive to noise,easy to detect false edges and miss the real edge.In this paper,an improved edge detection method based on Canny is proposed to deal with the log image,and the adaptive median filter is used to denoise.It avoids the difficulty of selecting the Gauss filter parameters and the high and low thresholds of the traditional Canny algorithm,as well as the phenomenon of the slow change edge loss.The eight direction gradient template is introduced to obtain the amplitude of the edge,which can be effectively detected.More edge details are measured;Otsu method is used to determine high and low threshold,avoiding the problem of artificial setting high and low threshold.Finally,the edge image is refined by morphological method.After getting the results of the log Canny edge detection,the logs are identified by the improved random Hough transformation method,and the actual area is calculated by the identified logs,and the number of logs,length diameter and the log volume are calculated.In this paper,the proposed algorithm is tested on the Matlab platform and compared with the traditional Canny edge detection results.It is proved that the proposed algorithm can effectively retain more edge details of the image,and suppress the interference of the false edge or the area with the similar color features.The edge of the log is extracted completely,which lays a good foundation for identifying the logs and calculating the log parameters after the random Hough transformation.By comparing the experimental results obtained in this paper with the actual log volume results,it is found that the relative error of the original wood accumulation theory measured by the algorithm is relative to the result of the actual measuring ruler by about 1.5%.Based on the improved Canny edge detection algorithm and the random Hough transform algorithm proposed in this paper,a log volume measurement software is implemented on the Android mobile platform.The main functions of the software are log image acquisition,log image mosaic,log image edge detection,log recognition and log parameter measurement.The software of the log volume measurement method proposed in this paper is easy to install on Android mobile phone and easy to operate.