Key points are not available for this paper at this time.
The technique of machine vision is extensively applied to agricultural science, and itshows great perspective especially in plant protection field.The paper describes a software prototype system for machine-vision intelligent decisionprecision sprayer. The final objective of this study is to set up a set of software system including twosub-systems. One is used to identify the image of cotton insect pests in field in real-timeautomatically, while the other to provide spray intelligent decision.The real-time digital images of cotton and insects captured by CCD were processed bycomputer, including image growing, image segmentation and features obtaining (i.e. color, shapeand texture etc.). The methods evolved in this system are both image processing and mathematicalmorphology. Subsequently, computer expert system will implement intelligent decision via artificialneural network and fuzzy logic method. The object of this system is cotton insect pests (Helicoverpaarmigera (Hubner) and red spider) in North China. Image database including cotton in differentperiod and different state is set up in this system, which is embedded in computer. The expertsystem determines the time, area and volume of spray pesticide. According to the result of imageautomatic identification and scientific decision, the system will achieve precision spray.The prototype system of automatic identification of cotton insect pests and intelligent decision(AICIPID) is under way. The results will be beneficial for reducing the volume of spray pesticide,enhancing benefits and protecting environment from pollution.
Li et al. (Wed,) studied this question.