Font Size: a A A

Research On Tobacco Disease Recognition Method Based On Image Processing

Posted on:2019-09-10Degree:MasterType:Thesis
Country:ChinaCandidate:Q ZengFull Text:PDF
GTID:2428330548484477Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
Tobacco products play an important role in the national economy.The yield and quality of tobacco leaves are affected by pests,diseases,growthing environment and planting methods.Their quality directly affects the health of consumers and the sales of tobacco products.The traditional identification method of tobacco leaf pests and diseases mainly depends on artificial discrimination and is subject to artificial and other subjective factors,which lead to the recognition efficiency and accuracy are not ideal.The rapid development of image processing technology provides a solid foundation for researchers to quickly and accurately identify and detect pests and diseases.This paper focusing on tobacco leaf diseases,makes some research about the key issues in tobacco leaf disease identification.The main work and innovations of the dissertation are as follows:(1)For the complex background of tobacco leaf diseases,a new region-based active contour model method is proposed.The new SPF function is used to evolve to reach the purpose of segmenting the leaf surface contour.Take the tobacco disease image,cucumber disease image and leaf image of vegetable as an example.The experimental results show that this method can well remove the complex background in the image,reduce the interference on the lesion segmentation,and obtain a relatively complete outline of the leaf surface.(2)In order to accurately segment the tobacco leaf disease of disease spot image and identify the types of disease,the method of preprocessing the image with Gaussian smoothing and Robert cross-gradient sharpening is proposed to enhance the boundary information and texture information of the leaf lesion image.In addition,the morphological segmentation method based on Lab color space is used to divide the leaf spot,and the results show that this method can well segment the lesions of tobacco leaves.Combining the shape features and texture features of tobacco leaves,related parameters are extracted to identify the types of tobacco leaf diseases.(3)The Tobacco Diseases and Insects Information Bank is designed to solve the problems of disorderly information,difficult to find and difficulty in data collection of tobacco leaf pests and diseases.This information library not only contains information related to tobacco leaf pests and diseases,but also has functions of managing databases and serving users.The establishment of the Tobacco Diseases and Insects Information database provides a better resource platform for agricultural workers and other researchers,meanwhile,it is convenient for users to search and download resources,as well as provide suggestions about the information in the database.
Keywords/Search Tags:Tobacco leaf disease, Active contour model, Lab color space, Information bank
PDF Full Text Request
Related items