Dissertation > Excellent graduate degree dissertation topics show

Research of the Real-time Face Detection Based on Video

Author: WangDaQing
Tutor: ZhangJianMing
School: Jiangsu University
Course: Applied Computer Technology
Keywords: Face Detection Adaboost Integral image Multi-pose Class Haar features Cluster analysis CAMSHIFT algorithm
CLC: TP391.41
Type: Master's thesis
Year: 2010
Downloads: 135
Quote: 4
Read: Download Dissertation

Abstract


Face detection is a very active computer vision and pattern recognition research topics have been widely used in video surveillance, human-computer interaction, image retrieval, video conferencing, authentication, virtual reality, and many other aspects. With the intelligent computing technology advances, new methods, new technology and constant introduction gives a face detection injected more vitality. In this paper, the comprehensive analysis of the previous face detection algorithm based on multi-pose face detection method based on AdaBoost and multiple decision tree, the following major elements: (1) In accordance with the rotation angle of the multi-pose face in training The sample angle space is automatically divided on the basis of multi-pose face detection method based on multiple decision trees. The method introduced in data mining FCM algorithm to automatically space is divided on the multi-gesture samples to solve the problem of the point of view of the multi-pose face samples identified uncertain. Guarantee the detection rate optimization algorithm the principle control sample space split, split the principles of computing by FCM clustering algorithm and the complexity of the algorithm to obtain the optimal classification. This method does not reduce the detection speed premise obtained detection tree having a greater discrimination performance. (2) in-depth study ADABOOST algorithm class Haar feature calculation cascade classifier, test results and treatment proposed adaptive step input image traversal methods for duplicate detection problem of face region. This method uses sliding down the mirror image of control traverse the pyramid structure of the image and the image detection window, take full advantage of the cascade classifier positive detection rate, to avoid unnecessary redundant computation, and omitted detected in the post-processing links, shorten the testing time, and improve the detection efficiency. (3) the use of the motion characteristics of the video of the human face, the existing target tracking algorithm based on the proposed method of automatically tracking face detection results. The way to avoid detection operations for each frame of the video, to save computing time and improve the real-time performance of the algorithm, and in accordance with the variation of the video human face tracking and monitoring, in the judgment of the tracking failure, automatically re-detects the face. (4) In order to verify the effectiveness and feasibility of this paper, the improved algorithm, this article will on multiple data sets detection experiments. Experimental results show that: the detection method solves the existing Adaboost face detection system in the multi-pose face detection failure. Detection accuracy for various unobstructed multi-pose face image, the detection system regardless of face detection rate or false positive rate of the non-face, are better than existing detection systems.

Related Dissertations

  1. Research on the Classification Based on the Reconstruction of Solder Joint,TP391.41
  2. Tongue Feature Extraction and Research of Fusion Classification,TP391.41
  3. The Fatigue State Recognition of the Driver Based on Eye Detection,TP391.41
  4. Feature Extraction, Selection and Combination in Lipreading,TP391.41
  5. The Research on Paper Currency Classification Method Based on Harr-Like Feature and Minimal Ball Including Samples,TP391.41
  6. Development of EST-SSR Primers and Application in Analysis of Genetic Realtionships in Tree Peony,S685.11
  7. Pre-hypertension syndrome characteristics,R259
  8. Comprehensive Quality Assessment of College Students,G645.5
  9. Studieson Effects of Soybean Species on Yuba and Initial Establishment of Quality Evalution System for Yuba,TS214.2
  10. ISSR Analysis of Genetic Diversity on 21 Lotus(Nelumbo Nucifera) Cultivars,S682.32
  11. Research on Predicting Intrinsic Disorder Protein Structure Based on Supervision Manifold Learning Algorithm,Q51
  12. Research on the Soil Environmental Function Zoning,X321
  13. Comparison of Gene Expression Data Cluster Methods and Gene Network Construction for Phytophthora Sojae Genes,S435.651
  14. Study on Heterosis and Genetic Basis of Soybean,S565.1
  15. Evaluation on Forage Quality and Biomass Energy Characters of Inbred Vegetative Lines of Napier Grass,S543.9
  16. Clustering Method Research Based on Divided and Conquered Method,TP311.13
  17. Elements of Thirty Kinds of Proprietary Chinese Medicines and Classification Based on Elements,R286.0
  18. Vehicle detection based on machine vision and vehicle distance measuring method,TP274
  19. Community-oriented education, personalized learning system and its implementation,TP391.6
  20. Design to E-learning System in Senior Vocational School Base on Moodle,TP311.52
  21. Face Recognition Method Based on DE,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile