With industrial development, as well as its impact on the climate, the haze has increasingly become a common weather phenomenon, and outdoor applications, the picture quality of the monitoring system to cause a great impact. At present, the actual project, mainly through other means, including radar, infrared sensing and monitoring of targets in the climate of fog and haze, the ideal response program in the field of video surveillance technology and cost constraints, radar, infrared and other not yet. Real-time video through the fog technology can bring a wide range of value for the video surveillance. Known through the fog algorithm can be roughly divided into two categories: a non-model image enhancement method, by enhancing the contrast of the image, to meet the requirements of subjective vision to achieve the purpose of clarity; other is a model-based imagerecovery method, which examines the reasons for image degradation, the degradation process will be modeled using the reverse processing, and ultimately solve the image restoration problem. Enhanced through fog to deal with the typical methods include: histogram equalization, filtering, transformation methods and methods based on fuzzy logic. The histogram equalization method, the global method of computing the amount of small but the details of the enhancement is not enough; partial equilibrium approach is better, but it may introduce a massive effect, calculate the large amount of noise is magnified and the algorithm is effective is not easy to control. Filtering through the fog transform algorithm, local treatment can get a relatively good outcome, but their calculation is a huge amount of resource consumption, is not suitable for real-time requirements of the equipment. Through the fog effect is not an ideal method based on fuzzy logic. Improve the image contrast can be enhanced to a certain extent, and by enhancing the region of interest to enhance the degree of identifiable. But the method is not the reasons of image degradation process from the start to compensate, so it can only improve the visual effect can not be good through the fog effect. Based on image restoration methods are mainly the following categories: filtering method, the maximum entropy method and image degradation function estimation method. Filtering methods such as Kalman filtering method, the whole large amount of calculation. The maximum entropy method to obtain a higher resolution nonlinear computation, numerical solution of the difficulties. Image degradation function estimation based mostly on the physical model (such as the model of atmospheric scattering and polarization characteristics through fog model) to design, capture multiple images at different time points as the reference image, in order to determine the physical model parameters , the ultimate solution to get the results of images in the fog-free state. This limits the application of such methods in real-time monitoring. security monitoring camera are now used in a variety of complex scenes, bad weather, all-weather real-time monitoring network camera portability and power, the treatment effect, deal with the self-adaptability, more stringent requirements. Good video through the fog should be in the atmospheric transmittance model based on integration of image enhancement and image restoration technology advantage, and thus able to obtain the ideal image and the actual works of reference.The advantages of real-time video through the fog and the other through the fog method is mainly reflected in the following areas: 1ã€Âthe ability through the fog. Real-time video through the fog accurate according to the different depth of field to remove the fog of the corresponding degree. Traditional image enhancement method may close range through the fog better vision is still the remnants of a lot of hazy mist; the accuracy of the element of choice-based method for image restoration through the fog effect, the ingredient of choice for accurate through the effect of fog processing better, such as the composition of select inaccurate may get bad results, the method through the fog performance of the algorithm is not stable enough. 2ã€Âpermeability. The real-time video through the fog after the image is very transparent, good contrast. Other methods because it is not an accurate depth of field is estimated remnants of the mist, through the fog after the image is not very transparent. 3ã€Âthe high degree of detail preservation. Real-time video through the fog with a special treatment to keep the details of the image through the fog can hide the original details retained after the fog and even some enhanced, which is the other method is difficult to achieve. 4ã€Âcolor saturation, the ability to restore. Real-time video through the fog does not change the hue of the image but simply to increase the saturation of its color, the other through the fog may tone distortion. 5ã€Âdoes not cause the image Pianan phenomenon. Real-time video through the fog contains the brightness to improve performance, dim the phenomenon of the resulting image does not appear. Other methods may lead to reduced contrast. 6ã€Âthe wide range of applications. Real-time video through the fog can also be used to the fog-free image processing to improve the original image's contrast and saturation, to improve the image of a sense of permeability, played the role to enhance the image visual quality. Real-time video through fog to enhance the quality of the video surveillance from multiple angles. First, it is a transparent fog technology can be used for aerosols lead to a variety of weather conditions through the fog to deal with; it is an enhancement algorithm can significantly enhance the contrast of the image, the image is transparent, clear; can significantly enhance the image details, the original hidden image detail to fully display; can increase the saturation of the image, so that the images colorful and lively, vivid images through fog treatment to keep accurate colors, natural appearance, and thus obtained good image quality and visual experience. Real-time video through the fog technology with strong engineering capabilities. Through fog technology applicable megapixels HD resolution, including the images; in a variety of resolutions can ensure real-time accurate through the fog to deal with; depending on the target scene in real time assessment of the current fog concentration and adaptive adjustment through intensity mist, fog conditions without fear of a scene change; through fog technology to arrive at an accurate depth of field, the non-uniform mist of the same scene can be accurately removed without residue through the fog defects. From the present test results, the technology in outdoor video surveillance system has good application prospects. Through the fog in real time with video compression, intelligent analysis technology for greater value. Due to the current mainstream video compression algorithms are lossy compression, will have low contrast image details of the damage, while the fog video is generally low contrast details below normal, compressed by the encoding are often ambiguous and can not be recovery. Can effectively enhance the image contrast and detail, to ensure that valuable information will not be lost by compression coding, and significantly improve the information on the effectiveness of real-time through the fog. Similar to real-time image processing through the fog, the analysis results of error rates, especially false negative rate can significantly reduce, thus greatly improving the usefulness of intelligent analysis system for the intelligent analysis. Scenes from the application point of view, the real-time video through the fog can be used for a variety of outdoor applications, such as the side of the highway traffic accident-prone areas, highway mount position; bus driver assistance facilities monitoring region; the focus of concern by the public security organsplaces and areas; power plants and power transmission equipment, key control areas; primary and secondary schools, the business center of the city and the city square. From the industry point of view of applications, including the transport sector, the public security industry, education, aviation, digital products, remote sensing image processing, food safety monitoring, and even military applications. Products and solutions from the application of real-time video through the fog in the field of security monitoring can be applied to the front of the camera, quick ball through the fog and enhance the quality of the image; can be applied to the DVR in order to enhance the quality of the image; can be applied to the large screen display to enhance the color saturation and image quality; embedded client software can also be used to enhance the quality of the preview image.
|