关键词: 叶面积指数, 半球摄影法, 无线传感网络, 定量遥感反演

Abstract: Aiming at the problems of large workload and low efficiency of groundbased measurements of LAI, a network observation system for LAI based on hemispheric photography was developed.Based on hemispherical photography and Wireless Sensor Network (WSN) technology, the system improved the image adaptive segmentation algorithm to accurately segment vegetation canopy images under complex environment, so as to obtain canopy structure information such as LAI in real time; for the issue of fluctuation of raw data at 5-minute collection frequency, the “most stable window” precision correction algorithm was proposed to extract daily LAI representative values. The results of comparison experiments with LAI-2200C showed that the two LAI measurements were highly significantly correlated, with the coefficients of 0.95, 0.69 and 0.87 for maize, woodland and grassland vegetation types, respectively. The correlation coefficients of the measurement results of the two instruments in the multivegetation type experiments in Liyang, Jiangsu and Changchun, Jilin also reached 85% and 81%, respectively, both of which were highly significant levels. The comparison experiment with MODIS MOD15A2H longterm series products also showed that the R2 of the two reached 098, which accurately reflected the dynamic change trend of wheat leaf area from green to maturity. The system has the advantages of fast and accurate, economical and convenient, wide range of use and multisite longterm monitoring. The instrument has been applied in many observatories, and its development will provide strong support to the remote sensing products authenticity validation and scientific research and practice in the fields of agriculture, ecology and climate change.

Key words: leaf area index, hemispherical photography, wireless sensor network, quantitative remote sensing inversion