tailieunhanh - Mobile Robots Perception & Navigation Part 6

Tham khảo tài liệu 'mobile robots perception & navigation part 6', kỹ thuật - công nghệ, cơ khí - chế tạo máy phục vụ nhu cầu học tập, nghiên cứu và làm việc hiệu quả | Contour Extraction and Compression-Selected Topics 191 where ne -number of even valued elements chains no --number of odd valued elements chains T- scale factor proportional to grid spacing. -Inverse chain The inverse chain of an 8-directional chain code may be obtained using the following relation c-1 Cj 4 where -addition mod 8 Example -For the curve shown a write the Freeman chain code using the 8-directional scheme. b Find the length of the chain code. c Find the inverse of the chain code. Fig. Chain encoding example a In the Figure below is shown the tracing of the curve using the 8-dierectional scheme. The chain code is obtained as 21010001212 b The length of the chain code assuming T 1 is L T ne n 2 1 7 4 2 c The inverse of the code is . Cị-1 Cị 4 65454445656 192 Mobile Robots Perception Navigation OCF and MSCE Methods of Contour Extraction The separation of objects from their background referred to as segmentation of gray scale images and as contour tracing or boundary following for binary images often provide important features in pattern recognition and scene analysis and is used in variety of other applications for example recognition of human faces from essential contours. In general contour extraction from two-dimensional images may be accomplished in two operations 1 Edge detection 2 Contour tracing. 1 Edge detection The aim of edge detection is to identify and enhance edges pixels belonging to boundaries of object of interest in the original image. An edge element is defined as a picture element which lies on the boundary between objects or regions of different intensities or gray levels. Many edge detection methods have been proposed for detecting and enhancing edges in digital images. Most of these methods are implemented as some form of gradient operators. Some images can be characterized as containing some objects of interest of reasonably uniform brightness placed against a background of differing brightness. Typical examples .