The Design Of The Corn Weed Identification System Based On Android Platform

With the popularity of mobile terminal further, led by Android operating system continues to mature, while the Android system derived program has become more abundant, endless, deepening and enriching people's daily life, agricultural applications will also continue to expand their product to mobile end, the Agricultural attract users but also greatly enhanced the increase in agricultural production, creation of income for farmers, this paper focuses on weed identification maize farmland, farmland due to weed competition between the often limited resources, Over time will interfere with the growth of corn, in order to get rid of the weeds cleared, they begin a lot of spraying herbicides. But it ignores the hazards of improper use of herbicides to humans, animals and the environment. From the practical application designed corn weed identification system based on Android platform. Operation of the system in reference to the results of previous studies on the basis of continuous improvement and the comparison test algorithms to identify the feasibility of the method is suitable for weed identification. Overview and analysis of relevant theory and image recognition technology Android platform, lay a theoretical basis for this study. Then based on the overall structure of maize weed Android platform identification system design. On the frame content includes many aspects of the overall structure of the system, software system design. In terms of content to study the inside of cornfield weeds image pre-processing algorithms, and its design. Used in the study design to the gray-scale image processing, multi-faceted approach thresholding, edge detection. Then design a corn weed image feature recognition algorithm based on ANN and extraction algorithms, including corn on the proposed image features and image features proposed weed, to differentiate between the two, in order to distinguish them in the corn weed. Finally, the paper design of corn Weed recognition system implementation and testing of the entire article summarizes research and further studies were analyzed.