本研究是針對鋪面維修專家系統以數位影像偵測方式，獲得之鋪面裂縫影像作為維修方式之推演時，由於一般繁忙道路或機場跑道進行鋪面調查及影像擷取的工作，多半僅能在夜晚進行，在光縓照射的不均勻情況下，所擷取的鋪面裂縫影像不足以是供專家系統作鋪面管理使用，故必須將影像照度不均勻進行補償動作；本研究使用三次多項式配合最小二乘法來補償光源反射不均問題，然後使用區段法（Segmentation）之模糊（Fuzzy Entropy）原理找出裂縫與背景之影像門檻值，然後進行二元化處理，最後使用物標示法（Object Labeling）方式將鋪面裂縫影像中的裂縫與以物件化，其目的可有效濾除影像上的斑點等非必要資訊，以提昇鋪面管理專家系統效能，最後並將上述的論理進行測試與驗證於既有之機場鋪面維修專家系統上。

For this study we used a CCD video camera to capture pavement images and segmentation in image processing to analysis distress images to improve the pavement management expert system. During investigation processing, the CCD video camera capturing pavement of distress images can not be used because the investigation work has to process during the nighttime and the illumination source is non-uniform that cause the images can not be provided the expert system to manage pavement. Therefore, before the images provide the expert system to manage pavement, the images should eliminate their non-uniform illumination source problem. After the CCD video camera captured the pavement images, we used three orders Polynomial function and Least-Squares method to eliminate images of gradation which have non-uniform surfaces due to the illumination at night. The Fuzzy Entropy method calculates image of threshold segments and creates binary images. Finally the Object Labeling algorithm finds objects that are cracks or noises from the binary image based on volume pixels of the objects. We sued these algorithms and tested them on the airport pavement management expert system, also providing some discussion and suggestions