Reliable and fast structure-oriented video noise estimation

Abstract

The purpose of this paper is to introduce a fast automated whitenoise estimation method which gives reliable estimates in images with smooth and textured areas. This method is a block-based method that takes image structure into account and uses a measure other than the variance to determine if a block is homogeneous. It uses no thresholds and automates the way that blockbased methods stop the averaging of block variances. The proposed method selects intensity-homogeneous blocks in an image by rejecting blocks ofstructure using a new structure analyzer. The analyzer used is based on high-pass operators and special masks for corners to allow implicit detection of structure and to stabilize the homogeneity estimation. For typical image quality (PSNR of 2040 dB) the proposed method outperforms other methods significantly and the worst-case estimation error is 3 dB which is suitable for real applications such as video suweillance or broadcasts. The method performs well even in images with few smooth areas and in highly noisy images.

An Improved Structure-Based Gaussian Noise Variance Estimation Method for Noisy Images

“…In order to alleviate the overestimation of the amount of noise in [13], the proposed method in this paper presents new masks for homogeneity measurement and a more accurate threshold for the averaging process.…”

“…In [13] the authors showed that the dividing pattern of 5 ×5 blocks is a good compromise between efficiency and effectiveness. Thus, the same dividing pattern is used in this paper.…”

“…Then the homogeneity for every block is checked. In [13], a uniformity analyzer based on 8 masks and high-pass operators was used to measure homogeneity, as is shown in Fig. 1.…”

“…Recently, a novel structure-oriented noise variance method has been proposed [13]. This method is also based on blocks, but it takes the image structure into account and uses a measure other than the variance to determine if a block is homogeneous.…”

“…When information on such noise becomes available, an image processing algorithm, such as filtering [1]- [3], edge detection [4]- [6], image segmentation [7]- [9], or image compression [10]- [12], can be adapted to the amount of noise to provide significantly improved performance [13]. Therefore, noise variance estimation has been a fundamental and important task in image processing systems.…”