An efficient background extraction and object segmentation algorithm for realtime applications

Real-time search-free multiple license plate recognition via likelihood estimation of saliency

“…Unlike [17] where the colour images are used, in our algorithm, the processing time is reduced when grey images are used. Figure 2 shows our proposed background extraction method.…”

“…On the other hand, the extracted background in [17] is fixed, whereas in our algorithm, the pixels values can be updated.…”

“…Finally, using CCA, the objects are detected. Comparing to our proposed object detection method, [17] proposes an algorithm of object segmentation based on background subtraction. In this method, the difference between the desired region and the background is calculated for both luminance and chrominance (signals used in video systems to express the colour information of the image).…”

“…In our proposed object detection method, the coordination of each object using the information of current and previous image frames is used, and similar to [17,18,19], the background extraction algorithm is performed. For images with dark shadows and irregularities in the shape, tracking algorithms such as [12,13,14,15] could also detect the objects.…”

“…Wang et al [17] proposed an effective real-time background extraction and moving object detection method with less memory usage comparing to the methods such as Chiu et al [18]. Jiang et al [19] proposed another background extraction algorithm, i.e.…”