Không gian kỹ thuật để phân loại ảnh (A) Một kết nối khu vực rộng lớn được tạo thành bởi pixels sáp nhập đã được gán nhãn là dữ liệu Mall đường phố ở DC (b) nhỏ gọn hơn Sau khi chia tách các tiểu vùng trong khu vực (a) (c) kết nối khu vực rộng lớn được tạo thành bởi pixels sáp nhập đã được gán nhãn như lát gạch trong Trung tâm dữ liệu (d) nhỏ gọn hơn Sau khi chia tách các tiểu vùng trong khu vực (c) | Spatial Techniques for Image Classification 239 a A large b More c A large connected region d More compact sub- connected compact sub- formed by merging pixels regions after splitting the region formed regions after labeled as tiled in Centre data region in c . by merging splitting the pixels labeled region in a as street in DC Mall data FIGURE See color insert following page 240. Examples for the region segmentation process. The iterative algorithm that uses mathematical morphology operators is used to split a large connected region into more compact subregions. region polygon boundaries. The statistical summary for a region is computed as the means and standard deviations of features of the pixels in that region. Multi-dimensional histograms also provide pixel feature distributions within individual regions. The shape properties 5 of a region correspond to its Aarea Orientation of the region s major axis with respect to the x axis Eccentricity ratio of the distance between the foci to the length of the major axis for example a circle is an ellipse with zero eccentricity Euler number 1 minus the number of holes in the region Solidity ratio of the area to the convex area Extent ratio of the area to the area of the bounding box Spatial variances along the x and y axes Spatial variances along the region s principal major and minor axes Resulting in a feature vector of length 10 2008 by Taylor Francis Group LLC 240 Image Processing for Remote Sensing Region Classification In the remote-sensing literature image classification is usually done by using pixel features as input to classifiers such as minimum distance maximum likelihood neural networks or decision trees. However large within-class variations and small between-class variations of these features at the pixel level and the lack of spatial information limit the accuracy of these classifiers. In this work we perform final classification using region-level information. To use the Bayesian classifiers .
