Research On The Algorithm And Real-time Implementation Of Infradred Image Enhancement

At present, the large-scale Infrared Focal Plane Array (IRFPA) with 512×512 detectors is one of the advanced infrared detectors in the world. Although some technological advances have been achieved in the IRFPA in recent years, Infrared images are created by temperature difference between targets and their background. Great bright background and low contrast are the primary characteristics of infrared images. These result the poor image quality and visual impression. So reducing the noises and enhancing contrast are the two important technology in infradred image processing.Based on the analysis to the noises forming mechanism and their characteristics, several different de-noise and enhancement algorithms are proposed to restrain noises influence and enhance the contrast of infradred images. For aming at the infradred images in real-time implementation, an adaptive image enhancement algorithm based on plateau histogram equalization are used in this thesis. By adaptively selecting the plateau threshold value ,this algorithm can enhance the targets and constrain the background noises. Moreover, based on the characteristic of output signal of IRFPA a 512×512 IRFPA imaging signal processing system is designed, which implement real-time procession to the IRFPA imaging signal, such as nonunifromity correction, blind pixels compensation and the infradred image enhancement. The experimental results validate the good effect in this systerm. Also, the system has the advantages of large dynamic-range, high procession-precision, stabilized performance, high speed, vigoroso flexibility, etc.