Pilot study on toward-target precision pesticide application in forestry

The paper introduced briefly the concepts of precision pesticide application and proposed the strategy of developing precision pesticide application in China, which will reduce the impact of pesticides on environments, increase the ecosystem sustainability, and protect the operators. The tree image processing, tree crown recognition and smart spray execution system were developed. The tree image processing system included a CCD camera, an image grabber, a computer, and an experimental set-up. The software was developed to realize the functions as follows: capturing, extracting and processing the key frame of tree image from video stream, segmenting the green target from its background by a classifier called relative color indices, measuring the segmented image by statistic method to obtain morphological and location features of the target, which provides into for intelligent understanding and pattern recognition. The tree crown recognition system based on BP neural networks was developed as the toward-target spraying pesticides depended greatly on the tree crown types. 6 typical tree crowns, that is, cone, oval, spherical, cylindrical, umbellate and truncate, could be determined. The spray execution system composed of spraying table, nozzles, solenoid valves and relays was developed, which will realize the data communications and convert spraying decision making results to instructions for controlling the spray execution system. The results of experiments showed that the developed systems were practical and suitable to recognize the tree images, the tree crown types. The non-target or off-target pesticide contamination could be decreased through precisely spraying the pesticide on the desired tree target.