| 3D film "Avatar" swept the world in2009, its stunning3D visual effects, realistic presence shocked the audience, changed people’s viewing. With a3D TV,3D digital home theater gradually into the tens of thousands of households, but also makes the industry "film into the3D era assertion. Existing3D video technology has been able to provide viewers with a better spot to experience3D audio technology is lagging behind the products currently on the market is limited to follow the original stereo or surround sound technology, these technologies can not meet consumer demand for3D audio sound effects needs. In order to achieve an immersive audio-visual experience, there must be three-dimensional sound field sound effects synchronized with3D video content, the internationally renowned institutions and large corporations to carry out the three-dimensional audio technology research, MPEG ISO as well as formulating three-dimensional Audio technical standards, the average home user expectations can be extended to three-dimensional audio technology, which makes the three-dimensional audio technology ushered in unprecedented opportunities for development, and become a hot research field of multimedia technology and an important development direction.Required in order to obtain a better three-dimensional sound, arranged in a large number of speakers, some systems can reach hundreds. The surge of the number of the channel such that the amount of data of the three-dimensional audio traditional audio several times, even several times, in the case of restrictions by the real-time broadcast bandwidth and storage media capacity, will reduce the reconstruction of three-dimensional audio effect. By parametric coding method to reduce the coding rate, the bit rate of the parameter will significantly increase with the growth of the number of channels, and the spatial parameters there are a lot of perceived redundancy, perceptual characteristics to remove the subjective redundancy of the parameter space parameters I can maximize the reduction of the parameters of the multi-channel bit rate. In addition, a large number of speakers arranged no uniform standards and principles of the arrangement, so that the various systems are not compatible, limiting the three-dimensional audio technology popularization and application. Speaker arranged to minimize human perception distortion in principle, by the human ear can guide the spatial perceptual characteristics speaker arrangement and reconstruction, reconstruction of audio and video listening experience. In summary, to carry out the study of the mechanism of three-dimensional space perception, the establishment of three-dimensional space binaural cues and location cues perceptual model, to provide theoretical support for the efficient coding and reconstruction of three-dimensional audio.In this thesis, the National Natural Science Foundation of China basic theory and key technology of mobile audio coding (No.60832002) and the basic theory of the three-dimensional audio coding and key technologies (No.61231015), the Ministry of Education, Fund for the Doctoral space perception The amount of information measure theory and algorithm research (No.20090141110054) project funded under the perceptual characteristics of the spatial cues and its application to study, from the three-dimensional space binaural cues and location cues perceived mechanism analysis, based on the perception of three-dimensional audio parametric coding of these three directions of innovative achievements.Binaural cues perceived characteristics of direction in three-dimensional space, research binaural line Suoqia perceptible difference between the parameter values and the frequency. The spatial audio coding method by extracting the Characterization of the spatial orientation of the binaural cues to the objective in addition to the inter-channel redundancy, but also the existence of subjective redundancy, and with a higher proportion of the increase in the number of channels parameter redundancy, even more than the the bit rate of the main channel, is still not effective the subjective redundancy removal method. To solve the above problem, this issue explores the spatial parameters to perceive the presence of redundant mechanism, the binaural cues perceived characteristics of traditional energy domain extended to the parameter field to obtain the relationship between binaural cues sensing threshold value and the frequency and binaural cues, perceptual model established by surface fitting binaural cues. This study from the frequency domain and the parameter domain binaural cues fine-grained perception experiments, making the model and the mechanism of human perception is more consistent, and can be represented by the mathematical method of binaural cues perceived variation for parameters the subjective redundancy removal of important guiding significance of this part of the research to the National Natural Science Foundation of China mobile audio codec basic theory.Location cues perceived characteristics of direction in three-dimensional space, location cues perceived characteristics analysis and modeling studies will be carried out three-dimensional space. The number of channels of the three-dimensional audio substantial increase in efficient data compression and reconstruction facing the challenge orientation perceived characteristics is an effective way to solve the above problem, existing research orientation-aware location-specific characteristics of qualitative analysis, pursuant to create space Orientation perceptual model will not be able to guide the three-dimensional audio compression and reconstruction. To solve the above problem, the subject will explore three-dimensional space leads to perceptual mechanism, test audio capture three-dimensional space in different locations through the unique design of the experimental setup, the establishment of the test source database, design, adaptive and psychology hearing test for the entire three-dimensional space locative clues sensing threshold value and, on this basis, to create a three-dimensional spatial location cues perceived sensitivity representation model. The results through the surface fitting to obtain the entire three-dimensional space of position clues perception threshold, provided the theoretical support based on the perception of three-dimensional coding and sound field reconstruction and other research.Perceptual characteristics based on the perception of the parametric3D audio encoding direction, the use of three-dimensional space binaural cues and location cues guiding space quantization and coding parameters. The existing three-dimensional parameters of the audio signal encoding, in the bit rate by the real-time broadcast bandwidth and storage media capacity limit, the parameter quantization errors will result in a sense of spatial orientation of the three-dimensional audio distortion, the three-dimensional space of the audio sound quality will be significantly decreased. Solve the above problem, the subject of the three-dimensional space binaural cues and location cues sensing mechanism is introduced to the parameter coding, lossless coding framework proposed perception of spatial parameters, the human ear can perceive only the quantization parameter change amount, removal of the perception of the parameters redundancy. Compared with the existing three-dimensional audio encoding method, binaural cue coding bit rate can be reduced to approximately15%, approximately25%of the orientation parameter coding rate.The research topics in the basic theory and key technology is expected to become a national and even international support3D audio standard technology, and enhance our core competitiveness in the high-speed growth of three-dimensional audio industry, the international competition for full participation in the field of three-dimensional audio and standardization work has laid a solid basis for further study. |
PDF Full Text Request