Abnormal color correction method applicable to thermal imaging video colorization

Abstract

The invention relates to an abnormal color correction method applicable to thermal imaging video colorization, comprising the following steps: (1) finding out changing pixels on a current frame relative to a previous frame; (2) respectively carrying out local region growing on the current frame by taking each changing pixel as an initial point; (3) working out the main color of each region, and judging colors of which the difference is greater than a threshold relative to the main color in the region as abnormal colors; and (4) using the main color of each region to correct the abnormal colors in the region. Compared with the prior art, the abnormal color correction method is low in complexity, can be used to detect and correct abnormal colors in real time in the process of colorization, and can significantly improve the accuracy of colorization results.

Description

technical field [0001] The invention relates to a method for correcting abnormal colors of color images, in particular to a method for correcting abnormal colors suitable for colorizing thermal image videos. Background technique [0002] Any object whose temperature is higher than absolute zero has thermal radiation, and the intensity of its radiation is related to the temperature of the object. The thermal imager detects the thermal radiation of the object, and forms a video image of the temperature distribution of the measured scene, that is, a thermal image. Thermal imaging is widely used in the field of night vision observation because of its high detection of thermal targets and high adaptability not affected by environmental conditions such as light and haze, and plays an important role in both military and civilian applications. [0003] From the perspective of imaging effect, thermal imaging technology for night vision has gone through three stages of development: g...

AI Technical Summary

Problems solved by technology

However, any method may inevitably produce local coloring deviation due to inaccurate estimation, inaccurate tracking or inaccurate matching during inter-frame processing, that is, abnormal color

Moreover, this abnormal color will gradually accumulate and expand with the recursive process until it significantly affects the effect of the entire picture, making the rendering of the video sequence unable to continue to be effective.

Images