Dissertation > Excellent graduate degree dissertation topics show
Based on high resolution satellite images of traffic flow parameters Extraction
Author: SunZuo
Tutor: LiangYanPing
School: Beijing Jiaotong University
Course: Transportation Planning and Management
Keywords: High-resolution satellite imagery Object-oriented image analysis Support vector machine classification Feature space Traffic flow parameter extraction
CLC: TP391.41
Type: Master's thesis
Year: 2010
Downloads: 232
Quote: 5
Read: Download Dissertation
Abstract
|
Existing traffic sensors collect traffic information there are many limitations , along with high-resolution satellite images commercialization emerges from a wide range of remote sensing images for road traffic information , has become the intelligent transportation system information collection technology , a research hotspot. This paper aims to study from high-resolution satellite images in the detection of vehicles, and further extracting traffic flow parameters. The study includes the following aspects: ( 1 ) characteristics for high resolution remote sensing images , drawing on visual interpretation of the principles of object-oriented classification of vehicles taken detection method . By multiscale segmentation and target feature extraction, creating an image object feature space to achieve the vehicle identification and classification ; ( 2 ) based on support vector machine classification method of detecting a vehicle . According to the characteristics of the vehicle in the image , the image texture feature extraction , image texture features obtained , from which the selected sample of vehicles and non- vehicle samples , training the classifier , set the kernel function parameters, image classification , enabling the vehicle to detect ; ( 3 ) results of data analysis for vehicle detection for traffic flow parameter extraction , including direct and indirect traffic flow parameters extracted traffic flow parameter extraction ; ( 4 ) QuickBird2 experimental data using two detection method is validated , the results show that the object-oriented image classification more accurate vehicle detection method , and the overall accuracy of the error matrix theory, more than 90% . And further select the two sets of data , the method comparison test , analyze the factors that affect the detection performance , including roads lined with trees and buildings is caused by dark shadows the main vehicle detection error . In this study, for intelligent transportation systems rich in information collection technology , traffic flow parameter extraction means expansion is important.
|
Related Dissertations
- Yeyahu typical wetland plant information extraction based on the analysis of spectral characteristics,P237
- Sports Metadata Extraction Technology Research and Implementation Based on Video Content,TP391.41
- Studies of Some Key Techniques in Hyperspectral Classification,P237
- Research on Video Based Face Tracking and Recognition,TP391.41
- Research on Face Recognition Approach Using Subspace with Embedding Neighborhood Discrimimant Relation,TP391.4
- The matrix perturbation a number of issues research,O151.21
- Application of Water-flooded Zone Identifying Based on Computational Learning,TP391.4
- Research on Technology of Human Face Detecting and Tracking,TP391.41
- Study of Water Quality Assessment and Parameter Prediction Based on Support Vector Machine,X824
- Compression encoding of the video image of the human face,TP391.41
- Applications of SVM to Predict Silicon Content in Hot Metal,TF54
- An Eigenspace Based Approach for 3D Object Recognition,TP391.41
- Projection on Speech Features Space Improves the Performance of Speaker Identification,TP391.42
- Medical Image Classification Based on SVM,TP391.41
- Study and Related Implementation of Robust Algorithm for Adaptive Beam-forming,TN911.7
- Continuous Audio Stream Segmentation and Classification Systems Research,TP18
- Method of Monitoring Vegetation Information on the Mining Based on Multiple Remote Sensing Data,P237
- Chinese Personal Name Disambiguation in Web People Search,TP391.1
- Multi-class Classification Algorithm Research Based on Fuzzy Support Vector Machines,TP181
- Support Vector Regression and Its Application,F224
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
|