A method of color space selection for color forest inspection image denoising applications

Abstract

The color forest dynamic inspection images corrupted by some noise are a large amount of original information obtained in the processing of defending the forest diseases and insect pests. Based on analyzing three color image attributes such as resolution, pixel depth and true color or pseudocolor, it is separately presented what are the external evaluation rules about the means and effect of the forest plant diseases and insect pests color image denoising processing in real time in RGB, HSI and YCbCr color space. The experiment results show that color space must be reasonly selected during the course of color image denoising. The advantage of applying HSI or YCbCr color space is that the denoising arithmetic executes faster and it fits the circumstance of less noise than applying RGB color space. The advantage of applying RGB color space is that the denoising arithmetic executes slower space and it fits the circumstance of more noise than applying HSI or YCbCr color.