ASABE Technical Library - Abstract

Abstract. In order to quickly and nondestructively monitor the crop growth status in the field, the canopy multispectral image of Winter Wheat In the field was acquired. The images were captured by the 2-CCD multi spectral image intelligent sensing system. The image sensing system was installed on the vehicle platform and moved at a vehicle speed of S (0.54m/s).Image fuzzy was produced due to the random relative motion. In this study, the point spread function was acquired through the spectrum estimation method. The blurred image was restored using Wiener filtering, constraint least square restoring method and blind deconvolution restoring method respectively. The quality of image judged by the image quality evaluation function show that, fuzzy image can be suppressed effectively by Wiener filtering method compared with constraint least square restoring method and blind deconvolution restoring method. Fuzzy image restoration effect is more satisfactory and it can provide technical support for the follow-up study. The canopy image of Winter Wheat in large field was pretreated by adaptive smoothing. Then, image was restored using Wiener filtering method and segmented based on HSI color model. Image parameters (Average gray value of R, G, B, NIR, NDVI, NDGI, RVI, and DVI) were extracted. The correlation between detection parameters and chlorophyll index was analyzed. By NDVI and RVI, the multiple linear regression (MLR) model of chlorophyll index was built with detecting R=0.64 and predicting r=0.63. The results show that: it provided support for the fast and non-destructive detection of winter wheat chlorophyll content based on vehicle platform.