| Winter wheat leaf chlorophyll content is an important factor affecting the growth of winter wheat. To solve the complexities, leaf damages and other issues caused by the chemical method for measuring the chlorophyll content of winter wheat leaves, and to broaden the scope of application of spectral measurement model, this paper taken the winter wheat planted in northwest arid and semi-arid regions as samples, has obtained the visible /near-infrared spectral images of wheat leaves using a high spectrometer, and studied the spectral analysis models of chlorophyll content of winter wheat. Contents of this paper include:(1) In order to cancel the spectral noise and to select the proper modeling method, the derivative, smoothing, multiplicative scatter correction, and standard normal variable conversion pretreatment for the samples have been employed, and then the chlorophyll relative measurements and the full wave modeling for the spectral reflectance of wheat leaves have been made with PLS(Partial least-squares, PLS). The modeling results have shown that the multiplicative scatter correction method is better than the other three methods in the chlorophyll content measurement of 100 samples. Compared with BPNN(Back Propagation Neural Network, BPNN) prediction models, the R and RMSE of PLS model is respectively0.8593 and 2.8115, and is better than the BPNN model.(2) To simplify the prediction model, the spectral data extraction feature bands for winter wheat leaves have been made by SPA(Successive Projections Algorithm, SPA), and 9 feature bands selected. On this basis, the SPA-PLS model has been made with the parameter R=0.8782, RMSE=1.5062, and the basic prediction requirement has been satisfied. Extracting two characteristic bands of 565 nm and 769 nm, 2-SPA-PLS model R is equal to 0.8363,RMSEP 1.7237. Compared to the SPA-PLS model, the modeling effect of 2-SPA-PLS has been slightly worse, but better than the full-band model. The experiments have shown that two band model can be applied to measure the chlorophyll content of winter wheat in northwest arid and semi-arid regions, and to simplify the hardware circuit design, and has great potential to practical application.(3) To effectively use the real-time and fast characteristics of spectral measurement process, the spectroscopy measurement system of chlorophyll content of winter wheat leaveshas been designed and implemented based on B/S structure. And the experimentation with the actual sample data has shown that the system is able to effectively complete data processing,and to meet real-time spectral analysis and chlorophyll content measurement of winter wheat leaves. |