Dissertation > Excellent graduate degree dissertation topics show

Study on Key Techniques of Content-Based Audio Retrieval (CBAR)

Author: PanWenJuan
Tutor: LiuZhiJing
School: Xi'an University of Electronic Science and Technology
Course: Applied Computer Technology
Keywords: Content - based audio retrieval Feature Extraction Audio Segmentation and Classification Example audio retrieval
CLC: TP391.3
Type: Master's thesis
Year: 2008
Downloads: 228
Quote: 1
Read: Download Dissertation


In recent years, with the rapid development of multimedia technology and network technology, increasingly rich information resources on the network, the information retrieval technologies produce a profound change. People are no longer satisfied with the traditional text-based retrieval, but need a video, image, audio and other media for fast retrieval engine. Content-based audio retrieval technology (Content-Based Audio Retrieval: CBAR) came into being. Extracted directly from the audio semantic clues retrieval based on semantic clues semantic retrieval process and media directly linked to make the search more effective, and adaptable. This paper first introduces to CBAR technical background and development process; then elaborated a successful and effective CBAR application of a variety of key technologies and the improvement of the existing audio segmentation classification algorithm; given retrieval experimental results and analysis; pointed out the deficiencies of the system and the future development direction. Sound and effective audio segmentation and classification is a prerequisite for retrieval by the system. Traditional the characteristic threshold-based segmentation and classification method using relatively simple features and previous experience handling the classification problem is relatively simple. Meanwhile, the feature threshold selection is also more difficult. In this paper, the Gaussian model-based segmentation algorithm, and given a new feature Mel-ICA improved the algorithm. The method does not require sample collection, to be divided according to the characteristics of change point, and achieved good segmentation results. This article also gives a combination classification method based on the threshold model combines the advantages of these two methods, while using the wavelet transform and Fourier transform to extract audio features to improve the accuracy of the classification. The characteristics and descriptions of the audio system the key. In this paper, using the time domain, frequency domain and time-frequency domain analysis methods to characterize the substance of the audio signal from a different angle, to constitute a description of the audio signal operator. Audio retrieval example the audio query mode (QBE), the first to use the minimum spanning tree (MST) clustering method to form a key frame, and then matched comparison of the same type of frame, reducing the strength of the calculation, greatly improving the retrieval efficiency. The experimental results show that the proposed method can better audio retrieval achieved better performance. This paper summarizes the work and proposed further research and exploration direction.

Related Dissertations

  1. Tobacco Diseases Auto-Recognition Research Based on Image Processing Technology,S435.72
  2. Research on Nondestructive Detection Technology for External Qualities of Papayas Based-on Vision,S667.9
  3. Research of Diagnosing Cucumber Diseases Based on Hyperspectral Imaging,S436.421
  4. Inspection and Recognition on a Class of Tool Processing Trace,TP391.41
  5. Research on Method of Grain Feature Extraction Based on FPGA,TP391.41
  6. Blends of different fibers and quantitative analysis of image recognition,TP391.41
  7. Based on sparse representation of high spatial resolution remote sensing image texture description Method,TP751
  8. Subject-oriented web -based technology research and implementation of classification,TP393.092
  9. Real-time video display system key technology research,TP391.41
  10. CT images of the lung nodule feature extraction method,TP391.41
  11. Term Weight-Based Chinese Text Classification Algorithm,TP391.1
  12. Automatic Classification of Pavement Based on Laser Data,TN247
  13. Research on Biological Visual Features Extraction in Images Combining Sparse Coding,TP391.41
  14. Research of Image Processing for Seam Track Based on Laser Optical Sensor with Structured Light,TP391.41
  15. Neural Signal Analysis and Decoding of Primary Motor Cortex in Rats,TN911.6
  16. Research of Preprocess Technique for 3D Space Handwriting Based on 3D Accelerometer,TP212
  17. The Research of Product Feature Extraction in B2C Website,TP311.13
  18. The Research on Fault Diagnosis Method Based on Continuous Hidden Markov Model with Feature Weighted,TH165.3
  19. Research on Feature Extraction of Cough Sound and Its Application in Identification,TN912.34
  20. Research of Sleep Stage Based on ECG Signal during Day Time Nap Sleep,TN911.7
  21. Extraction of Pleural Indentation and Lung Cavity of Lung Nodules in CT Images,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Retrieval machine
© 2012 www.DissertationTopic.Net  Mobile