基于矛盾解决理论与物质——场分析的实木地板节子识别新方法研究

针对传统区域生长方法识别实木地板节子存在准确率低且速度慢的问题，运用TRIZ中矛盾解决理论分析与物质——场分析，提出一种结合分水岭、区域生长以及边缘检测的新的实木地板节子识别算法。算法首先将原图像转换为灰度图像；其次，运用形态学分水岭的方法对灰度图像进行分割；再次，选取满足条件的种子区域进行区域生长，得到节子区域；最后，运用Sobel算子对图像进行梯度运算，并找到节子的边缘。仿真实验表明，该算法较传统方法能够找到更合适的种子区域和区域生长的阈值，实现了对节子的快速、完整提取，节子分割平均用时60ms，平均辨识准确率在90％以上。

A New Method for Solid Wood Floor Knot Recognition Based on Contradiction Solution Theory and Substance-Field Analysis

In view of traditional region growing method for solid wood floor knot recognition has low accuracy and slow speed problems, in use of contradiction solution theory and substance-field analysis in TRIZ to analyze, a solid wood floor knot recognition method combined with watershed, region growing and edge detection was proposed. The method converted the original images to gray image at first, and then completed gray image segmentation by morphological watershed. Next, the method selected seed regions meet the conditions for region growing, got knot region. At last, knot region was found through the use of sobel operators for the gradient operation to the image. It is observed that this method can find more suitable seed regions and regional growth threshold than the traditional methods, and extract the knot completely and accurately. The time used for knot segmentation was 60ms respectively, and the average accuracy rate of identification is above 90%.