Restricted by camera hardware, digital images captured by digital cameras are noisy, and the noise content of each color channel of the digital image is not balanced. However, most of the existing denoising algorithms assume that the entire image noise is constant, causing errors in the denoising of color images (non-uniform noise images), affecting image noise removal and texture detail protection. To solve this problem, we propose an evaluation operator that can describe the noise content and texture content in the local area of the image. According to the description value, the image pixels are classified, and heuristic denoising parameters are selected for each class to achieve a balance between noise removal effect and texture retention effect. Experimental results of multiple denoising methods show that the proposed algorithm has better denoising effect on color images.