| The plant is an important component of the biosphere, and plays an irreplaceable role in our life. The classification of Botany is a very important foundational work of the Botany research and the forestry production. This work mainly based on plant's exterior characteristic. Leaf is a very direct-viewing key character of the plant, different plant may be different in leaf color, the foliate structure as well as the leaf vein distribution, and this enables us to use the plant leaf to carry on the classification of the plant. In the past, this kind of work was by completes artificially, not only consuming much time, the work load was also very heavy, moreover the working efficiency was relatively quite low. Along with the current pattern recognition technology and the computer imagery processing technology's fast development, we can use the computer to help us with this work. So we can reduce the work load greatly, enhance the analysis accuracy and promote the working efficiency.This article has conducted the massive paper search to the pattern recognition technology and the digital image processing technology, and then introduced the LBP method. The LBP method is an effective method which carries on the image texture analysis and has so many advantages. The method has a widespread application in many front edges. This article has carried on the outline of the LBP method's related concept and the theory, and then proposed an innovative application of the LBP method in the plant classification recognition domain.The main work of this article includes: use digital image processing technology to pretreat the leaf sample image; study and propose the theories of applying the LBP method on the feature extraction of plant's leaves; programme each kind of the LBP operators; design and realize the leaf recognition system based on the LBP method.Finally carried on the experiment test to this system, then did some detailed analysis to the result and carried on the forecast to the later work. |
PDF Full Text Request