Method and device for reducing intraframe noise

Abstract

The invention discloses a method for reducing intraframe noise, comprising the procedures as follows: Reading the K-1 rows pixel of the current frame and the current K rows pixel are obtained with the arrival of the K-th pixel; taking the currently arrived pixel row as the current pixel row and filtering the intermediate pixel point in each pixel block with size KXL in the acquired current K rows pixel; judgign whether the current pixel row is the last pixel row of the current frame. If not, updating the pixel row in the acquired current K rows pixel the rows difference between which and the current pixel row is K-1 by utilizing the current pixel row. The current K rows pixel are obtained with the arrival of the next pixel row of the current frame and returned to execute the filter operation; otherwise, the procedure is finished. The invention has the advantage that the invention discloses a device of reducing intraframe noise simultaneously, which can eliminate intraframe noise effectively under low-light level and improve images quality.

Description

technical field [0001] The present invention relates to image processing technology, in particular to a method and device for reducing intra-frame noise. Background technique [0002] In a low-light environment, the line noise or point noise output by the sensor (Sensor) is very large. In order to reduce the noise as much as possible and preserve the details of the original image, various methods have been proposed at present. [0003] Specifically, a relatively simple method is to use a low-pass filter (LPF) method. The LPF is a device used to transmit only signals with frequency components lower than a reference. That is to say, when an image signal with a frequency change is input to the LPF, and the amplitude of the output signal is observed, a characteristic is obtained, that is, the level of the high-frequency component is relatively low. [0004] However, when observing from different points, LPF uses the average value of the observed pixel and its surrounding neigh...

AI Technical Summary

Problems solved by technology

However, using this method, that is, the LPF uses the surrounding pixels as an average value, it will cause the information on the edge of the image to be reduced similarly to noise

In this way, although the noise is reduced, the entire image is blurred, which in turn affects the image quality

Images