鐵路平交道移動障礙物機器視覺辨識研究

Study of Moving Obstacle Detection at Railway Crossing by Machine Vision

陳立偉(Li-Wei Chen)

指導教授 ： 蒲永仁

長榮大學/健康科學學院/職業安全與衛生學系(所)/碩士(2010年)

https://doi.org/10.6833/CJCU.2010.00173

摘要

在台灣鐵路交通運輸是相當重要且融入於民眾生活環境中，鐵路與公路交會之處「平交道」正是用路人與火車交會之區域，火車帶來了方便，卻也造成了許多的交通意外事故。93-97年間，平均3~6天就會發生一起事故，這數據可發現國人沒有確實遵守交通規則外，更突顯平交道的安全防護措施仍有待改善。
現有的平交道防護設備包括平交道標誌、閃紅燈、警示鈴、遮斷器，以及上下行列車指示號誌，上述安全措施只能警告用路人，而無法防止各種闖入與臨停的事件發生。近年來裝設的緊急按鈕，常因緊張或惡作劇，降低了緊急按鈕的功效。國外已發展出的主動式偵測設備包括：紅外線光電、感應線圈式、廣域雷達以及超音波偵測裝置主動偵測障礙物，但這些安全系統都有使用上之缺點，例如：偵測死角、風勢、氣候、環境干擾，以及法令管制等影響。
本研究期望提供平交道安全備援系統另一途徑，利用平交道縮小模型，搭配LabVIEW圖控式程式語言及IEEE1394 CCD攝影機擷取鐵路平交道之模擬影像，進行鐵路平交道移動障礙物機器視覺安全系統之研發，藉由影像辨識程式，偵測出移動障礙物，為了達到上述目的，本研究發展之軟體先將灰階影像轉為黑白，經由設計之實驗找出最適化二值化門檻，當設定下限值30上限值230時，畫面雜訊與前景較為適合後續影像處理的狀態，並利用型態處理實驗選定侵蝕一次及膨脹四次，作為型態處理之組合；在以上參數與不同亮度條件下進行驗證發現，平均警報準確率為97.6%，平均誤差時間約為0.157秒，辨識正確率Recall平均為98.1%，以及Precision平均為99.5%。由於使用此程式辨識平交道移動障礙物，警報準確率與辨識的正確率之失誤機率甚小，能提供平交道安全有效防護，未來將能夠對火車及時發出警報，降低事故的機率及嚴重性，以維護車廂內司機員、乘客與道路使用人在平交道上行的安全。

關鍵字

鐵路平交道 ； IEEE1394 CCD攝影機 ； 機器視覺

並列摘要

The railway transportation is quite important, and has integrated into the populace's lives in Taiwan. The crossing where a railway and a road intersect is the region that the road users and the trains meet. The railway has brought us convenience, but on the contrary has also caused many traffic accidents each year. According to the statistics, an accident would take place every 3~6 days at all crossings in Taiwan during 2004-2008, which shows that some citizens do not obey the traffic regulations regularly and that the safety protection measures on the crossings still need to be improved.
The existing protection devices on a crossing include those traditional signs, flash lights, horns, barriers and direction indicators of the approaching trains. These devices only give warmings to the road users but can not do anything about intrusions and temporary stops at the crossings. Among the recently equipped devices emergency buttons need to be operated manually. People’s nerves or hoaxes may actually reduces the effect of this device. Other recently developed devices can actively detect obstacles, which include photoelectric sensors, inductive coils, radar and ultrasonic systems. However, these safety devices have some weaknesses, such as the blind spots, the influences by wind and climate, the surrounding interferences and the restrictive regulations, etc.
This study is designed to be a redundant safety measure which is able to detect the existence of obstacles at the crossing. We use a IEEE1394 CCD camera and the LabVIEW Graphic programming language to process the images of a railway crossing miniature. To achieve this, we first determined by experiments the optimal range of the binarizing thresholds from 30 to 230, which turns the grayscale images into the black and white ones. In so doing, both noises and foregrounds are then to be processed by the basic morphological transformations. The experimental results suggested that one erosion together with 4 dilations are suitable for the foreground recognition. According to the above paremeters and the various brightness conditions, the evaluations of the developed software demonstrated that the average alarm accuracy reaches 97.6%, the average time of error is around 0.157 sec, the average Recall from the accuracy of recognition is 98.1%, as well as the Precision is 99.5% averagely. Due to the high accuracies to recognize the moving obstacles, this software is expected to provide effective protections for the railway crossings. Once the safety system is completed and able to send alarms immediately to avoid the accidents or reduce their severities, the safety of the crew and passengers on the trains as well as the road users can be further assured.

並列關鍵字

Railway Crossing ； IEEE1394 CCD Camera ； Machine Vision

參考文獻

[20] 姚錫齡，“平交道設置安全預警系統芻議＂，營建知訊，251期，pp.52-57，2003年12月。

[27] 陳彥良，“即時立體視覺物體追蹤系統＂，中原大學機械工程學系碩士論文，2003年7月。

[29] 陳頌婷，“鐵路平交道淨空機器視覺安全系統研發＂，長榮大學職業安全與衛生學系碩士論文，2008年7月。

[2] D. D. Giusto, F. Massidda, and C. Perra, “A fast algorithm for video segmentation and object tracking, ” 14th International Conference on Digital Signal Processing, vol.2, pp. 697-700, 2002.

[3] S. P. Lohmeier, R. Rajaraman, and V. C. Ramasami, “Development of an ultra-wideband radar system for vehicle detection at railway crossings, ” IEEE International Conference on Geoscience and Remote Sensing Symposium, vol.6, pp. 3692-3694, Jun.2002.