Research On The Color-Cast Correction Of Crop Digital Image

High-quality digital image is not only a prerequisite for information extraction, but also a guarantee for estimation model and physic-chemical parameters statistics established on qualitative or quantitative parameters. In this paper, take canopy and leaf image of rice as the main object, we try to do some research on influencing factors of image quality, the evaluation index of image quality, the correction of color cast image base on computer vision technology and using image processing technology. The main contents of this paper include:(1) The influence of rice cultivars, shooting angle, growth period, light angle, and the brand of digital camera on image quality under the different nitrogen levels. The results show that rice cultivars and the brand of digital camera have less effect on color information; there is no difference of processing results between images captured under 30°and 60; the correlation between color characteristic and SPAD values is inconsistent at different growth periods; the images captured along the light, its correlation between color characteristic and SPAD values is very well.(2) The article discusses the correlation between color characteristic and SPAD values. The correlation between (G-B)/(G+B), (G-B)/(R+G+B) and SPAD values is significant at different growth period, the correlation coefficients are 0.66 and 0.68. Not only the correlation can reflect an actual relationship between leaf colors and nitrogen contents, but also it can be used as digital index of quantity of nitrogen fertilizer.(3) This studies take white balance,the biggest value of color and gray card color correction methods to correct rice canopy and leaf image. The results show that the white balance color correction method is useless for rice canopy image, the effect of the biggest value of color correction method is general, the demand for actual scene in gray card color correction method is low, besides this, the method is simple, fast, and stable.To facilitate computer analysis of above issues, the digital image processing system for the field of agriculture was developed by our research groups. It can extract the morphology, color and textures of leaves through image analysis. I get some correlated analysis data and conclusions based on this platform.