Fusion rules and image enhancement of unmanned aerial vehicle remote sensing imagery for ecological canal data extraction.

For the problem of rough irrigation habits and low irrigation efficiency in Hetao irrigation area in China, which only pays attention to increasing yield, this paper proposes a method of accurate extraction of canal information based on fusion of rules and image enhancement in order to improve the level of canal irrigation management in Hetao irrigation area. Accurate calibration of canal information is the premise of precision irrigation management. For satellite remote sensing, it is difficult to collect information with high accuracy in sub lateral ditches in the field. For ground stations, calibration work is difficult to move flexibly with the target. Therefore, based on UAV remote sensing images and object-oriented segmentation method, a fusion of rules of two spectral features and two shape features was adopted. The optimal combination rule of "image enhancement splicing threshold 2%, normalized chromatic aberration coefficient a = 0.1, b = 0.9, c = 1.2, spectral average less than 98, minimum enclosure rectangle aspect ratio between its minimum value and 1, extension line greater than 1 meter" was obtained. The information of some ecological canal system in Hetao Irrigation District of Inner Mongolia was extracted, and the recognition accuracy reached sub lateral ditches level. The two evaluation methods of canal system calibration were put forward and the experiment was evaluated. The correct rate of sample 1 interpretation of combination rules was 87.5%, SNR (Signal-to-Noise Ratio) of classification results was 7.602, which provided accurate canal system information for precise irrigation operation and management. [ABSTRACT FROM AUTHOR]