A Method for Detecting Breaches and New Objects in Multiple Outdoor Images

This paper presents a new automated change detection method for detecting breaches in the integrity of and attached objects in front of fence wires in multiple outdoor images of the same scene containing fence wires acquired by a mobile camera from slightly different viewing positions, angles and at different times. To detect significant changes, edges of fence wires firstly have to be extracted from multiple outdoor images using a combination of the Sobel detector and an adaptive thresholding technique. Secondly, morphological operations such as dilation and erosion are applied into binary images produced by the previous process in enhancing the binary images. Next, an area-based algorithm is applied to enhanced binary images in separating small and big objects based on their average areas determined once in the calibration process. Finally, objects that survive are then fed into a fuzzy inference system in calculating their possibility values. Based on these possibility values, the survived objects can be classified as significant or unimportant changes. Experimental results demonstrate that the method has a high success rate (94.12%) in detecting true positives in these kinds of multiple outdoor images