Characteristic extraction, classification and identification algorithm based on two kinds of eaglewood micrographs

In order to classify the eaglewood, the work proposed a method of wood fiber segmentation and characteristic extraction based on the eaglewood micrographs. The active contour model was used to extract the contours of the eaglewood micrographs. After screening of wood fiber, the geometric features and shape factors were extracted to form characteristic vectors. After that, SVM was used to achieve the classification of the two kinds of eaglewood micrographs. Results show that the extracted characteristics can be used to express eaglewood fiber. With good classification performance, classification algorithm is easy to achieve.