Research On Fog Weather Image Clarity

Vision is the highest organ of perception, so that image plays an important role in human perception, and image processing is an important means of human visual continuity. Image enhancement and image restoration are two basic elements of digital image processing technology, the aim is to improve the image quality, in order to identify the human eye or computer processing and analysis. Images in the mist, rain and other natural weather conditions often get blurred, low visibility due to atmospheric particles so captured images appear distorted color, low contrast and other degradation, of course, which led directly to the role of outdoor imaging systems the partial loss to people's production and life brought some adverse effects. Therefore, in order to improve the function of weather conditions outdoors haze imaging system, effectively improve the degraded image quality in fog, fog on the image quickly and efficiently to deal with fog, try to restore the image of the details of the scene, with theoretical and practical urgent need.Firstly, the traditional Retinex algorithm SSR and MSR, dark colors priori, histogram equalization, wavelet transform these types of fog image processing algorithms integrated simulation analysis, comparative advantages and disadvantages of each method, preferably in combination knowledge of the causes of fog image processing method, and the results have learned to discuss research; Secondly, an improved image to fog Retinex algorithm, the first image processing MSR, then a power of transformation, a greater degree of image dynamic range compression to suppress highlight areas for emerging non-linear transformation of the last Wiener filtering to remove noise and enhance image detail, clarity of the image obtained. Experimental results show that the algorithm can effectively improve the fog captured image clarity, more suitable for human visual perception, better recover minutiae scene image, has good recovery results; the final study based on DSP chip fog days image sharpening system, which is divided into three parts:image acquisition, image processing and image display portion. Tools (1) in the image acquisition section for acquiring image fog is an analog camera. It can output in real time PAL system (referred to as the P format) or NTSC (the N standard) analog video signal, we use the N-standard CCD camera. (2) in the image processing section, using the image processor (DSP chip) is TI's TMS320DM642. First, the analog video signal analog camera collected by TVP5150 analog to digital conversion, the input video stream into 8-bit video data ITU-R BT.656 format, TMS320DM642 video processor through the mouth piece on the BT.656 data stream into the FIFO, and then through the EDMA channel data FIFO in carry-on-chip or external memory chip, then the caller image data processing to clear the fog, foggy day in order to achieve the purpose of image contrast enhancement. Then, the digital image data is processed through the encoder chip SAA7121 digital to analog conversion, and the converted image BT.656 data real-time display and output. Because encoding and decoding chip SAA7121 chip TVP5150 equipment different addresses, you can also share DM642 I2C bus. (3) the display portion of the image output, connect the data stream encoded output signal from the display, so that will be able to deal with; and the completion of image acquisition, processing and display of work by the device driver model DSP/BIOS real-time operating system It displayed clear images.Through the study of several traditional paper defogging algorithm, an improved Retinex fog image sharpening algorithms and constructed foggy image sharpening system, to achieve a great degree of validity foggy image clarity and practicality!...