剛性鋪面版塊影像破損比對處理系統之初步開發研究

Study of the preliminary development of the rigidity pavement broken slab image matching process system

陳俊穎(Jun-Ying Chen)

指導教授 ： 李釗

國立中央大學/工學院/土木工程學系/碩士(2006年)

摘要

本研究以Borland C++ builder軟體為工具，設計一系列版塊影像比對處理之程式模組，並嘗試就版塊影像所包含的相關資訊，如雜訊、接縫位置、裂縫區域及版塊破壞等情形，進行版塊模擬測試。版塊影像比對處理主要有四個步驟，首先為自動擷取版塊影像，其次為影像正規化，再來為版塊影像前處理過程，最後為影像比對部份。
影像比對處理係針對同樣位置版塊影像上，模擬明顯雜訊及含裂縫情況對版塊影像所造成的變異進行探討。研究結果發現，執行版塊影像比對處理前，必須先經合適的影像前處理過程，而影像前處理首重於最佳二值化影像，經最佳二值化影像經由四種邊緣偵測法處理後，研究發現Canny對於局部細微特徵點偵測具較佳處理成效。而版塊影像比對處理成果顯示，藉影像直方圖研判版塊影像內部資訊的成效良好，尤其以像素點累積量之增長情形可得知影像中可能含裂縫或破壞區域，亦可從區塊影像平均灰度值差異上，研判其版塊影像中區塊的破壞程度、破壞位置及破壞率。
利用本研究初步開發之剛性鋪面版塊影像處理比對系統，所擷取之完整破壞版塊影像，可供後續研究者進行版塊破損影像辨識處理分析，期以日後建立完整有效率之自動化鋪面破壞評估系統。

關鍵字

影像處理 ； 影像直方圖 ； 邊緣偵測 ； 影像比對

並列摘要

In our research, a series of slab images matching’s program model was designed by the software of Borland C ++ builder. We also try to implement the slab simulate test for the related information of the slab image. It contains the noise, seam position, crack area, and destruct situation of the slab, etc. There are four steps in the process of the slab image matching. The four setups are the automatic catching pavement images, image normalization, the slab image preprocessing, and the image matching, respectively.
The image matching which is discussed on the same position of the slab image simulates the obvious noise and the variation of the slab image by crack situation. In our research, a suitable image preprocess is needed before the proceeding of the slab image matching. An optimum binary image is very important for image preprocessing. Among the optimum binary image through four kinds of edge detections, the method of canny edge detection has better effect for fine features. Therefore, the inner information of the slab image which has better effect can be found by the image histogram. In this image histogram, the possible crack or the destructive area inside the image can be observed by some growth situations of pixels accumulating quantity, especially. The destructive degree, destructive positions, and the destructive rate of the block inside the slab image can also be observed by difference of the average gray value on the block images.
This research which utilizes preliminary development of the rigidity pavement slab image matching systems can carry out recognizing the broken slab images by extracting the complete broken slab image. It will be an efficient and complete automatic estimative pavement broken images systems in the future.

並列關鍵字

image matching ； image histogram ； Image processing ； edge detection

參考文獻

14.陳同孝、黎廣福、謝鑫棟，「植基於邊緣影像處理之影像比對搜尋引擎」，科技與管理學術研討會論文集，2000，pp. 395-402。

34.陳冠宇、張永鵬、傅培耕，「即時影像追蹤系統之設計與實現」，中原學報，Vol. 31, No. 1, 2003, pp. 21-29.

11.C. K. Leung, F. K. Lam, “Maximum Segmented Image Information Thresholding”, GRAPHICAL MODELS AND IMAGE PROCESSING Vol. 60, No. 1, January, 1998, pp. 57-76.

4.Tamikawa T, “A Study of road crack detection by the Meta-Genetic Algorithm”, Proc. of KSEEE ’99 Int’l. Conf., IEEE, 1999, pp. 543-548.

8.F. Y. Shih, C. F. Chuang, “Automatic extraction of head and face boundaries and facial features”, Information Sciences 158, 2004, pp. 117-130.