Information Extraction of Ecological Canal System Based on UAV Remote Sensing Data for Precision Irrigation

In view of the problem of extensive irrigation in water diversion irrigation in Hetao irrigation area, and the problem of increasing the planting area in order to increase the output, it is necessary that building a water-saving ecological irrigation area in Hetao irrigation area. The information extraction of canal system is the precondition of precision irrigation. Compared with satellite remote sensing, unmanned aerial vehicle (UAV) remote sensing platform is easier to improve resolution and more flexible than ground station. Based on the UAV remote sensing image, the image is preprocessed with ENVI 5.1 software, and the three channels of the remote sensing image are stretched and displayed respectively, and the contrast is enhanced by combination of the three channels. In this paper, the object-oriented method is used to segment the image, and the rule-based classification method is used. Based on the separate and combined analysis of different rules, the optimal combination rule of the spectral mean value is less than 98, the minimum bounding rectangle length width ratio is between the minimum and 0.85, and the lengthening line is more than 1 m. The recognition accuracy reaches to the sublateral ditches grade ecological canal and evaluates the extraction results. The accuracy of the combined interpretation is 96.4%, which provides the information of the canal system for the management of precision irrigation.