Infrared image cross grain removing processing method and processing device and storage device

Abstract

The invention discloses an infrared image cross grain removing processing method and processing device, and a storage device. The infrared image cross grain removing processing method comprises the following steps: acquiring an airspace cross grain value of a current frame airspace basic image of a current frame infrared image, and subtracting the airspace cross grain value of the current frame airspace basic image from a pixel value of the current frame airspace basic image to obtain a current frame airspace cross grain-removed basic image; obtaining a current frame time domain basic image and a previous frame time domain basic image, and obtaining a time domain cross grain value of the current frame time domain basic image according to the current frame time domain basic image and the previous frame time domain basic image; and performing anti-pollution processing and dynamic and static weight processing, and subtracting the anti-pollution processing and dynamic and static weight processing time domain cross grain value of the current frame of time domain base image from the pixel value of the current frame of space domain cross grain-removed base image to obtain an output imageof the current frame of infrared image. According to the scheme, the output image can have good definition and has no smear.

Description

technical field [0001] The present application relates to the technical field of thermal imaging, in particular to a method for removing stripes on an infrared image, a processing device and a storage device. Background technique [0002] Infrared sensors can convert infrared light into electrical signals and are the core components of infrared imaging equipment. Infrared sensors based on infrared focal plane arrays are the mainstream development direction of infrared sensors because of their small size, low cost, and high sensitivity. However, due to the current technological level, the output circuit of the infrared focal plane array usually shares the same output circuit with the same row (or column) of pixels, and the bias voltage of the output circuit is not exactly the same, so that different sensors on the infrared focal plane array are opposite to each other. The same infrared radiation produces different output signals, resulting in a streaky non-uniform noise in t...

AI Technical Summary

Problems solved by technology

However, due to the current technological level, the output circuit of the infrared focal plane array usually shares the same output circuit with pixels in the same row (or column), and the bias voltage of the output circuit is not exactly the same, so that different sensors on the infrared focal plane array are opposite to each other. The same infrared radiation will produce different output signals, making a kind of streaky non-uniform noise in the infrared image

[0003] In the existing design, the calibration-based correction method cannot update the parameters in real time, and it needs to assume that the parameters remain unchanged for a long period of time between two consecutive calibrations, while the bias voltage of the output circuit of the infrared focal plane array changes rapidly. Therefore, noise cannot be effectively eliminated; although the method based on scene correction can update parameters in real time, it takes a long time to converge, and the use of long-term image sequences may cause "smearing" phenomenon, which cannot effectively eliminate noise; frequency domain filtering is used method, it is effective for periodic stripes, but it cannot filter random stripes, and it is not easy to implement in hardware; in addition, the mean value of the current row is estimated by using the mean variance of adjacent rows, and then the original mean value of the current row is replaced by the estimated row mean value. This method It is only effective for images with a small range of temperature changes in the scene. For images with large temperature differences, the error in estimating the row mean with the mean variance of adjacent rows is large, and additional stripes will be introduced instead.

Images