Cotton Starscream Pest Detection Technology Research Based On Android

Cotton has always been an significant economic crop in our country’s agriculture. As one of the pillar industries of agriculture, the cotton planting industry has promoted the development of China’s textile industry and the promotion of the national economy. Cotton plant diseases and insect pests affected the yield and quality of cotton,according to statistics every year because of cotton pests caused yield losses caused by about 15%-20%, serious can reach 50% of the losses, which greatly affected the farmers’ income and the development of cotton industry. The traditional method of cotton disease and pest diagnosis is mainly based on the experience of artificial judgement, so the subjectivity is strong, the error is big.Farmers can not timely respond to treatment, resulting in less cotton production. Therefore, this paper analyzed the means the current detection technology of plant diseases and insect pests,combined with image processing and Android system, the detection system of cotton red spider insect pests based on Android was designed.Cotton red spider can cause serious harm to cotton yield cut more than 30%, in this paper, we design a kind of pest detection system of cotton red spider based on Android. The system had good real-time performance, low cost and portability good advantages, this paper mainly on the system structure and function, system process, software design, detection methods were studied.The following work mainly:(1)The analysis based on Android cotton red spider mite pest detection system overall structure and function demand, according to the system design principles for the selection of mobile terminal platform,based on Android cotton red spider mite pest detection system of overall and diagnosis of process design.(2)Selected OpenCV computer vision library to finish cotton red spider pest image preprocessing,feature extraction,recognition and diagnosis, image processing included gray image, image binarization,image smoothing, image segmentation and image compression, feature extraction involves the extraction of the edge, shape, and color of the cotton leaf affected by the pest infestation of the red spider. After according to shape and color features recognited cotton red spider insect pests. Finally, according to the ratio of pixel method were the diagnosis extend of damage of cotton red spider insect pests.(3)Built the development environment of Android system, used Android application basic components and application resources to complete the design of the user interface, cotton red spider insect pests image acquisition, feature of the main interface to browse and use, cotton red spider insect pests related information query and system help guide.(4)Server complete account of the basic operation, used OpenCV computer vision library java interface to calling image processing algorithms, dealt with the image of cotton red spider insect pests, the results of identification and detection displayed in the mobile phone terminal, finished Android cotton red spider mite pest detection system design, after tested the program, realized the detection function of cotton red spider insect pests.