Dissertation > Excellent graduate degree dissertation topics show
Research on Methods for Pedestrian Association Across Non-overlapping Camera Views
Author: LiBo
Tutor: LiGuoHui
School: National University of Defense Science and Technology
Course: Systems Engineering
Keywords: Non-overlapping cameras Target detection Feature matching Target association Feature Fusion
CLC: TP391.41
Type: Master's thesis
Year: 2011
Downloads: 15
Quote: 0
Read: Download Dissertation
Abstract
|
The target tracking non-overlapping monitoring system is an intelligent monitoring system is an important research direction , and it refers to the integration of multi- surveillance cameras to track the trajectory of the target . Goals related issues in a non - overlapping surveillance cameras is a core issue for target tracking in intelligent monitoring system . This paper studies the target associated with the non - overlapping surveillance cameras , and conducted in-depth research goals associated characteristics extraction and matching . Target associated technology status , analyze its current problems and a breakthrough design the necessary pre-processing and post-processing of the target associated with the matching process , and a goal-based appearance features describing the model matching method . The main content of the paper include : 1 . Still camera moving target detection , based on background subtraction methods designed based color clustering model of the target body area to extract the basic framework by introducing color segmentation method solved the problem of a target detection in the method of the background subtraction incomplete , and improve the effect of the target feature extraction . Target feature extraction and matching step for disadvantage of traditional MCSHR method , proposed distribution model based on color space , the better to deal with the appearance of the differences between the different goals . Meanwhile, the reconstruction of the target based on texture pattern matching method , more in line with the characteristics of human cognition , and design a specific matching algorithm can effectively improve the accuracy of target matching , rich characterization of the target appearance . 3 . Traditional goals associated method , did not make good use of the shortcomings of monitoring the video continuous image of the target characteristics , based on Adaboost the continuous target image sequence target associated methods . The use of a continuous target image feature training Adaboost combination of classifier, and for surveillance cameras in the non-overlapping target association , improving the accuracy of the target associated . Combination of the above methods , through experiments associated with moving target surveillance video . Choose a different sight surveillance video moving target feature extraction to match , and the associated results verify the rationality and effectiveness of the proposed method .
|
Related Dissertations
- Research on Joint Target Detection for Dual-Sensor Image and System Implementation,TP391.41
- Feature Extraction, Selection and Combination in Lipreading,TP391.41
- Moving target trajectory analysis based Intelligent Traffic Monitoring System,TP277
- The Study of Moving Object Detection and Tracking Algorithm Based on Image Information,TP391.41
- Research on Key Technologies of Monocular Visual Odometry Based on DSP,TP391.41
- Research on Moving Object Detected Technique for PTZ Cameras,TP391.41
- 360 degree rotation based on single- camera tracking system,TP391.41
- Monitoring image ROI extraction and target detection technology,TP391.41
- Distributed target detection in wireless sensor network research,TP212.9
- Face recognition algorithm based on feature fusion research,TP391.41
- The Compression and Fusion Technique Research of Underwater Target Feature,TN911.7
- Research on Two-dimensional Feature Extraction Methods,TP391.41
- Research on Color Face Recognition Based on Fusion of Components Feature Extracted by Spectroface and FLD,TP391.41
- Human Emotion Recognition Using Fused 2D-Fractional Fourier Transform Features Based on CCA,TP391.41
- Research on Moving Object Detection and Tracking Algorithm Based on Local Feature Analysis,TP391.41
- X-ray Medical Images Stitching Based on Corner Matching Using Two-way Asynchronous Voting Strategy,TP391.41
- Based on face and fingerprint multi- modal identification system research,TP391.41
- Camera-based network monitoring system,TP277
- Moving Target Detection Research on Key Issues,TP391.41
- Research on Multiple Objects Tracking Based on Detection in Video Sequence,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
|