|Title:||Automatic information summary and extraction from construction video monitoring systems|
|Subject:||Hong Kong Polytechnic University -- Dissertations|
Digital video -- Data processing
Building -- Superintendence -- Data processing
|Department:||Department of Civil and Environmental Engineering|
|Pages:||xiv, 228 pages : color illustrations|
|Abstract:||Automatic information summary and extraction systems are developed in this research for effectively processing and using video monitoring data at construction sites. The outputs of the systems are anticipated to assist construction managers in improving jobsite management, especially in improving workface productivity. The research addresses four critical issues related to the effective use of the enormous, dynamic, and unstructured construction videos. Advanced computer vision techniques are used and/or developed to address these issues. The major research components include (1) automatic video summary based on key frame extraction, (2) automatic macro progress measurement based on image registration, (3) automatic micro progress measurement based on work quantity measurement, and (4) automatic activity sampling based on trajectory classification and modeling. Video summary techniques are used to extract essential construction information. On-site video recording systems are increasingly used for monitoring construction activities. The recorded videos contain rich and useful jobsite information that can be used for a variety of purposes. A large amount of video data generated by continuous monitoring, however, creates tremendous challenges on data storage and retrieval. Due to the relatively slow pace of construction activities, a significant portion of the recorded data is redundant. Therefore, archiving raw construction videos into a concise and structured set of key frames would facilitate data storage, retrieval and analysis. Three key issues in automatic key frame extraction from construction videos are studied, including the selection of proper video features, scene segmentation, and key frame extraction. New image features and methods are developed to address the three issues. A validation experiment indicates that the developed features and methods can effectively and efficiently extract representative key frames from the complex and dynamic construction videos. Construction progress measurement is a key component in evaluating construction productivity. Progresses made at a large scale, namely macro project progress, can be obtained by comparing the video images taken at different times. Spatial data registration is a critically important step for image data comparison. In spatial data registration, a transform model needs to be fitted. In this study, three modifications are made to the Progressive Sample Consensus (PROSAC) method for such purpose, including tentative correspondence refinement, modification of progressive sampling and local optimization. The modifications are made to address the specific characteristics of image data taken from buildings under construction. Comprehensive comparisons and validations are made to verify the effectiveness of the developed modifications, and satisfactory results are obtained.|
The third research component is focused on progress measurement at workface, i.e., micro project progress. The micro project progress can be obtained by tracking and counting the completed work at the operation level. An image processing procedure is developed, with efficient computer vision methods proposed for each step in the procedure. Algorithms are developed for automatic calculation of two commonly measured quantities at workface, including planar area and patch quantity of completed work. The validation experiment shows that the proposed framework for planar area and patch quantity detection is feasible. Besides progress measurements, construction activity tracking is another important component in construction productivity evaluation and analysis. This research component is focused on activity tracking of construction workers or construction equipment, automatic classification of work activities, and assessment of work efficiency. In the proposed framework, a suitable moving object detection method is used to detect moving objects in an image frame. Object tracking methods are subsequently used to establish the correspondences of moving objects across different frames. The time utilization of each work cycle of a worker can be calculated by using single object tracking method. The efficiency or inefficiency of a worker can be determined based on the results of trajectory analysis in automatic activity sampling. The work activities are classified and modeled based on trajectories of involved workers on a construction site. Space and time utilization are analyzed using trajectories in multiple object tracking methods.
Files in This Item:
|991022164554803411.pdf||For PolyU Staff & Students||6.66 MB||Adobe PDF||View/Open|
|32282.pdf||For All Users (Non-printable)||6.7 MB||Adobe PDF||View/Open|
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item: