| Infrared polarization images have good penetration ability under complex weather conditions such as fog,which make them can identify targets in complex backgrounds.Therefore,in recent years,they are more and more used in military and other fields for target identification and detection.Traditional polarization images use pseudo-color processing algorithm,which has high complexity and large computation,resulting in slow processing speed and difficulty to complete real-time image processing.To solve the above problems,this paper completes the real-time processing of infrared polarized images from the aspects of simplifying the pseudo-color processing algorithm and speeding up the operation speed of the pseudo-color processing system.Based on the principle of polarized light generated by reflection and refraction of light,a pseudo-color algorithm is proposed that maps P and S polarization state image information and polarization degree information to the RGB color space after processing,which does not require complex calculations and greatly simplifies the processing process compared with the traditional polarized image pseudo-color processing algorithm.The pseudo-color effect maps processed using this algorithm show that different targets under test have different polarization information due to differences in material,roughness and other properties,which results in different colors being displayed in the pseudo-color effect maps.This makes it easier to distinguish different targets after pseudo-color processing,which is more conducive to target identification and detection.The TMS320DM642 chip is used as the core of the infrared polarization pseudo-color signal processing system.The signal processing system directly processes the original digital video signal from the detector,which reduces one video coding and decoding process and speeds up the processing speed compared to the general processing of the coded analogue video signal.The processing effect of the pseudo-color algorithm proposed in this paper and the processing speed of the designed system are evaluated.A comparison of the standard deviation,mean gradient and information entropy of the images before and after the pseudo-color processing shows that the pseudo-color processing scheme proposed in this paper has the effect of obtaining more useful information,increasing the image edges and enhancing the image quality.The real-time performance of the designed signal processing system is demonstrated by using the method of adding timer interrupts to the program. |
