Research Based On The Human Visual System Evaluation Model Of The Video

Along with the rapid development of computer hardware and software and thecorresponding solutions for the communications technology, China has proposed a tripleplay of strategic decision-making, and announced in this year’s Mobile World Congress onthe lead into the4G era, that a growing number of communications datarely on the videoimage form, this trend will inevitably lead to people will greatly enhance the video imageprocessing and transmission performance requirements.However, it is well known that thevideo images from the collection to the final presentation in front of the observer to gothrough a lot of data collection, processing, transmission and recovery, will inevitablyintroduce many different types of distortion in this tedious process. These distortions in theperception of the human visual system are the image quality changes. As the recipient of theimage information is true, so the ultimate goal of a variety of image processing technologyand the corresponding system are expected to get the best visual effect. The video imagequality is used to evaluate whether the image of the test images after treatment systemtechnical indicators to achieve the desired effect. Therefore a clear need for often in thecollection, processing and transmission of video image data information process through theestablishment of video quality evaluation model for video quality evaluation, in order tooptimize the system to achieve with the least resources to provide the best quality videoimages.At this stage of the video image quality assessment method has two major categoriesof objective evaluation and subjective evaluation. The traditional objective qualityassessment methods, such as peak signal to noise ratio and mean square error method ismore mature, but in many cases, these methods did not take into account the human visualcharacteristics, and its evaluation are often not able to meet the person’s subjective feelings.We all know that the information obtained from nature,80percent from the visual, thevisual has enabled man to perceive and understand the world around, it is the mostimportant human feeling. Because of this very necessary to study closer to the objectiveevaluation of the subjective feelings of the human visual.On the other hand, the subjectivequality evaluation method is well in line with the subjective feelings of mankind, but the evaluation of time-consuming, laborious and, more importantly, can not be embedded to thepractical application of the system, we have been clear from past research under standing ofthe objective both image quality assessment method has its own pros and cons, whichrequires reasonable optimization in future research work. So, the purpose and focus of ourresearch is to find an effective method has the advantage of subjective and objectiveevaluation of the largest to play out of the image quality assessment method, the same time,people began to think about two reasonable evaluation method integrated in a evaluationsystem will be more in line with the evaluation model of the subjective feelings ofhumanity.Method based on the characteristics of the human visual system to bring newinspiration to us to proceed from the parameters affecting image quality, the video qualityevaluation model proposed by some of the characteristics of the analog of the human eye,video evaluation of the noise situation. For both the physiological and psychologicalcharacteristics of the human visual system, a visual model can be used for video qualityevaluation. This model uses the gradient of the image frame to extract the internalinformation of the image, and with reference to the human visual system for the perceptionof several motion vectors in the visual scene modeling to simulate the observedcharacteristics of the human visual system moving objects. Finally, through the objectiveand subjective evaluation experiment has finally been able to better meet the humansubjective feelings of the video image quality assessment method.