Quantitative Analysis And Assessment Research On Factors Influencing The Visual Comfort Of Stereoscopic Images

Recently, with the rapid development of 3D technology and the popularization of 3D products, the stereoscopic display technology has won more and more favors through its distinctive visual feelings and immersive experience. It has a bright future in applications, such as industry and commerce, science, military, medical, education, entertainment, and so on. Among all kinds of stereoscopic display methods, the polarization is easier to operate that it has been widely used. Yet, because of the restrictions of viewing vision and angle, the viewers will feel uncomfortable if they spend a lot of time viewing the stereoscopic images without resting, which directly related to the visual security and physical health of the viewers, directly restricts the development and popularization of the stereoscopic display products. However, most of the current researches on factors influencing the visual comfort of stereoscopic images are qualitative, in order to judge the visual comfort of the given stereoscopic images more directly and accurately, to obtain standardized and unified criteria, it has a great significance to find the comfort areas of visual comfort and to study the influencing factors synthetically.The paper summarized the current researches on the visual comfort of stereoscopic imaging at home and abroad. Based on the human visual system and the stereoscopic imaging technology, it mainly concerned on the quantitative research of the contrast parameter, On the one hand, the paper built the database of stereoscopic images by the linear adjustment method and researched how the contrast parameter influenced the comfort of stereoscopic images by the subjective assessment method creatively. We got the comfortable contrast matching and difference maps, and verified the universal applicability of both two maps. The accurate rate of the two maps can reach more than 90%. On the other hand, the paper firstly applied the BP neural network algorithm to construct the comprehensive model of the different parameters which can judge whether one stereoscopic image is comfortable, then we verified the universal applicability of the BP neural network model and calculate the weight of four important parameters and analysis the influence separately.The experimental results show that, the subjective assessment can reflect the real feelings of the viewers and the neural network matches the regular of human's brain. So the model matches the subjective assessment greatly. Through the quantitative research and the comprehensive model, we can judge the visual comfort degree of a stereoscopic image quickly and accurately. The research can provide a base for the assessment of binocular stereo pairs and promote the development of stereoscopic display technology healthily.