Research And Implementation Of Image Analysis Methods In Plant Leaves

Agriculture is the foundation of our national economy, and the improvement in the efficiency of agricultural production and the extent of agricultural automation are the bases of modem agriculture. Transforming tradition agriculture using the rapid development the image processing technology, enhancing the technology content in the agricultural production, arming the traditional agriculture with the modern science and technology knowledge, already was the more and more agricultural workers' mutual recognition. This paper makes a study on plant leaf, analyze the image of leaf using the digital image processing technology, obtain leaf shape parameter, which we can use to analyze crops' growth situation and instruct agricultural production.This paper proposes the methods of preprocessing, following the contour of the leaf and computation of the leaf parameters, and has done the thorough research to the related image processing technology.In the image preprocessing stage, studying on the existing image enhancement algorithms, we use threshold-recursive median filter to smooth the image. The filter not only reduces time of computing, but also gets obvious better results. Then we adopt threshold processing to segment leaf image, and improve the quality of the binary image with mathematical morphology method.In this paper, we propose an improved algorithm of following contour based on connected component labeling, which can reduce the times of scanning image and accurately obtain outside contour of the leaf.On the bases of the result of contour following and connected component labeling, we bring on some methods to get parameters of the leaf. The experiment proves that these methods are effective. Chosing several parameters as the characteristic parameter, we carry on the simple classdified recognition of leaf-shape. This paper designs and realizes an analysis system of plant leaf.