Research On Plant Leaf Recognition Algorithm Based On Image Analysis

Protecting plant requires plant classification and recognition, and the plant recognition has practical significance. Plant leaves are the most important identification organs. Recognizing the plant leaves based on image analysis technology has a promoting role in popularizes the knowledge of plant classification and protection. To extract leaves features and recognize leaves types by image analysis, are the research concentration and main content of the paper. This paper surveys the image analysis based plant leaf recognition technologies in recent years, and divides the present methods into three categories which are based on relational structure matching recognition, based on statistical recognition and based on machine learning recognition. To improve the design of classifier and training time, a new method combining the clonal selection algorithm and K nearest neighbor (CSA+KNN) is proposed. Firstly, Maximum Class Square Error (OTSU) and Median filtering method are used to preprocess the image. Secondly, it gets plant leave geometric features through the Boundary searching method, the Vertex of convex hull decision algorithm based on Bi-polar angle and the Calculation of minimum enclosing rectangle. And it can get the fractal features through blanket algorithm and other texture features based on Gray level co-occurrence matrix method. Having the image preprocessing and getting the comprehensive features information from geometry and texture feature, the CSA+KNN is used to train and test plant leaf samples. The plant leaf database which has100leaf species is applied to test the proposed algorithm. Compared with other methods, the experimental results demonstrate the efficiency, accuracy and higher training rate of the proposed method, and verify the significance of texture features in leaf recognition.