基于数字图像处理的樟子松锯材分级研究

为了解决传统锯材检测方法效率低、劳动强度高和人为因素影响大的问题,笔者运用数字图像处理的方法对樟子松锯材等级进行检测,实现樟子松锯材等级检测的标准化和程序化。根据樟子松锯材缺陷与天然木纹颜色存在显著差距的特点,利用Otsu's方法的最佳全局阈值处理锯材图像得到最佳分割阈值,获得锯材的二值图像,得到缺陷位置像素坐标,将缺陷坐标与锯材实际尺寸对应,计算出相应锯材的净划材尺寸和出材率,完成对锯材等级的划分。利用该方法对75片试件进行表面质量检测和等级划分的准确率为94.67%,分级程序的运行时间约为1.793 s。该方法在保证检测准确率的情况下不仅可以减少检测时间、降低人工劳动强度、提高检测准确率,而且解决了人工检测主观性强的问题。

Classification of Pinus sylvestris Sawn Timber Grade Based on Digital Image Processing

In order to solve the problem of low efficiency, high labor intensity and human factors caused by traditional wood detection methods, in this paper, the digital image processing method is used to detect the grade of Pinus sylvestris sawn timber, and the standardization and programization of Pinus sylvestris sawn timber grade detection are realized. According to the significant difference between the sawn timber defects and the natural wood grain color, the sawmill image is processed using the optimal global threshold of the Otsu method to obtain an optimal segmentation threshold, a binary image of the sawn timber is obtained, the coordinates of the defect location are obtained, and the net material size and yield of the sawn timber are calculated, to classify the grade of sawn timber. The accuracy of surface quality inspection and grade division of 75 specimens by this method is 94.67%, and the running time of the classification program is about 1.793 seconds.The method described herein ensures the accuracy of the test. Not only can reduce the detection time, but also reduce the labor intensity, improve the detection accuracy, and solve the problem of strong subjectivity of manual detection.