The Research On Stereo Perception Quality Assessment Methods And Stereo Visual Characteristics

In recent years,three-dimensional image processing technology develops rapidly,and it has been widely applied to military,industry,education,medication and so on,which has changed people's life style.Stereoscopic image processing technology has become one of the hottest issues in signal processing area.People hope to obtain good image quality,however,stereo image inevitably be affected by noise from stereo acquiring process or stereo display process,which affects stereo perception.The research on stereo perception quality assessment method has become a challenging innovation research problem.Stereo camera shooting technology is the first factor affecting stereo image quality,and the noise from other process will further affect the stereo perception.To ensure the perception quality,whether stereo acquiring or stereo image display requires stereo image quality assessment,which will fill the vacancy of stereo perception quality assessment method in stereo technology area.So this thesis focuses on discussing stereo camera shooting quality,stereoscopic image quality and human stereo visual characteristics.Firstly,to solve the bad stereo perception problems caused by stereo capture technology,this thesis,based on stereoscopic image acquisition technology,analyzes the shooting rules under different shooting conditions,and proposes a comprehensive stereo camera shooting quality assessment method.Experimental results demonstrate that the proposed approach can effectively evaluate the shooting quality of stereo cameras.Secondly,based on two human stereo visual combination mechanisms in visual area:single channel fusion model and two channels fusion model,this thesis proposes two stereoscopic images quality assessment models: Stereoscopic image quality assessment method based on signal fusion model and Stereoscopic image quality assessment method based on two channels fusion model.Both of them are effectively applied the absolute disparity,which greatly reduces the algorithm complexity and saves the assessment time.Experimental results indicate that human stereo fusion characteristics play an important role in stereoscopic image quality assessment,and the above two models can be effective applied to evaluate the quality of stereoscopic images.Since the researches on single channel fusion model is not only more mature than that on two-channels fusion model,but also the evaluate model based on single channel fusion model takes human stereo visual saliency characteristic into consideration.So the overall performance of the proposed model based on single channel fusion model is better than that based on two-channel fusion model.Finally,based on the wide and effective application of human stereo visual characteristics in the field of stereo shooting and stereoscopic images quality assessment,this thesis studies human stereo contrast sensitivity characteristics,and builds the corresponding models,which pave the way for the application of stereo technology.