Читать книгу Handbook of Intelligent Computing and Optimization for Sustainable Development - Группа авторов - Страница 79

The GMG background subtraction model, proposed by Godbehere et al. [18], was used to obtain the dynamic foreground from a given input frame. We briefly describe this model as illustrated in Figure 3.2.

An input frame I(k) is quantized in color space and compared against the static background image model, Ĥ(k), to generate a posterior probability image. The resulting image is filtered using morphological operations. The filtered image is then segmented into a set of bounding boxes, , using connected components. The Kalman-filter bank maintains a set of tracked foreground objects, , and a set of predicted bounding boxes, , for time k. The Gale-Shapley matching algorithm pairs elements of with , and then, these pairs are used to update the Kalman-filter bank. The final result is the collection of foreground pixels, which is represented as .