工地安全監控系統

Construction Site Surveillance System

陳韋成(Wei-Cheng Chen)

指導教授 ： 丁肇隆

共同指導教授 ： 張瑞益

國立臺灣大學/工學院/工程科學及海洋工程學研究所/碩士(2015年)

https://doi.org/10.6342/NTU.2015.01490

摘要

近年來工地事故頻傳，營造業之重大職業災害發生比率，相較於其他行業更為嚴重。其中，工地人員裝備安全配戴不齊為造成嚴重傷害的主要原因。為了避免悲劇一再發生，藉由人體偵測及物件辨識技術，判定勞工進入工地時是否按規定著裝，以達到更全面性的安全檢查。
本論文主要分為三大部分：影像前處理、特徵擷取及辨識。首先，以網路攝影機拍攝影像後，透過背景相減法擷取移動的前景影像，藉由前景高度動態定位出頭部及軀幹位置，接著再各別抽取色調直方圖、飽和直方圖以及區域二元圖樣(Local Binary Pattern, LBP)作為特徵，再交由支持向量機(Support Vector Machine, SVM)進行分類。實驗結果顯示，本系統可以有效的辨識工地安全帽及工地背心，其準確率分別為97%和93%。

關鍵字

服裝辨識 ； 安全帽偵測 ； 背心偵測 ； 人體比例分析

並列摘要

Numerous construction site accidents have happened around the world in recently years. According to the Ministry of Labor, the occurrence rate of severe occupational injury in construction industry is much higher than others. This high risk is primarily caused by the deficiency of the personal protective equipment (PPE). In this thesis, we apply to the technique of body detection and object recognition on PPE checking system to examine whether construction workers are equipped as prescribed or not. As the result, the rate of construction hazard could be reduced.

There are three parts in our system which including image preprocessing, feature extraction and recognition. First, videos of workers are taken by an IP camera. Then, the moving foreground images would be extracted by background subtraction, and the positions of head and body are located by the height of the foreground image. Lastly, the Support vector machine (SVM) is utilized to perform classification on the features which are hue histogram, saturation histogram and local binary pattern (LBP). The experiment results show the system could effectively recognize the safety hats and safety vests with the accuracies of 97% and 93%, respectively.

並列關鍵字

clothing recognition ； hard hat detection ； vest detection ； torso proportions analysis

參考文獻

[3] M.-Y. Ku, C.-C. Chiu, H.-T. Chen, and S.-H. Hong, "Visual motorcycle detection and tracking algorithms," WSEAS Transaction on electronics, pp. 121-131, 2008.

[6] J. Chiverton, "Helmet presence classification with motorcycle detection and tracking," Intelligent Transport Systems, IET, vol. 6, pp. 259-269, 2012.

[10] Z. Zivkovic, "Improved adaptive Gaussian mixture model for background subtraction," in Pattern Recognition, 2004. ICPR 2004. Proceedings of the 17th International Conference on, 2004, pp. 28-31.

[11] S. Du, M. Shehata, and W. Badawy, "Hard hat detection in video sequences based on face features, motion and color information," in Computer Research and Development (ICCRD), 2011 3rd International Conference on, 2011, pp. 25-29.

[12] R. Waranusast, N. Bundon, V. Timtong, C. Tangnoi, and P. Pattanathaburt, "Machine vision techniques for motorcycle safety helmet detection," in IVCNZ, 2013, pp. 35-40.