Using SPOT5 fusion-ready image to identify the statistical characteristics of vegetation patches

Using high spatial resolution remote sensing image to study landscape structure characters is one of the important aspects in landscape ecology, and visual interpretation of image is often used, which needs lot of time and labours. In this paper, firstly vegetation patches are identified using the Canny edge detector and ellipticity, then the statistical characteristics of the vegetation patches are calculated. The experiments show that the detection accuracy of the patches is about 82.1%, and improved to 93.6% after the disconnection of the connective patches. The directions of about 70% patches are approximately south-north, which indicates that the approximately south-north is the main distribution direction of vegetation patches in the study image. The minimum and maximum area of the patches is about 50 m2 and 1206.25 m2, and the areas of about 60% patches are from 115 m2 to 415 m2. The histogram and cumulative histogram of the minimum distances between the patches show that the minimum and maximum value of the minimum distances among the patches is about 14.7 m and 102.8 m, and the minimum distances of about 91% of the minimum distances among the patches are from 12.5 m to 57.5 m. The rose graph indicates that the main azimuth between one patch and another patch which have the minimum distance is north-northwest, and the second is northeast and southeast, which reflects the distribution characteristics of the patches in the image. The presented study indicates that the integration the Canny edge detector with the algorithms for extracting circle and ellipse object based on ellipticity are simple and effective for detecting vegetation patches, and are easy to identify the statistical characteristics of the vegetation patches, which reduce the time and labours for the visual interpretation of vegetation patch.