Multi-stroke relaxation matching method for handwritten chinese character recognition

Cascade Markov random fields for stroke extraction of Chinese characters

“…In the past decades, the BU and the TD approaches have been two major strategies for stroke extraction from Chinese characters [17,10,31,22,8,4,11,3,15,28,23,19,14,2].…”

“…In contrast, the TD approaches use the character representation learned from examples to guide the stroke extraction process (for example, model-based stroke extraction [17,10,31,32] and multi-stroke relaxation matching [3]). Existing TD approaches need over-segmented pieces of substrokes (for example, line approximation method [17,10,3] and curvature detection method [31,32]), and involve an exhaustive search of almost all possible concatenations of substrokes.…”

“…Existing TD approaches need over-segmented pieces of substrokes (for example, line approximation method [17,10,3] and curvature detection method [31,32]), and involve an exhaustive search of almost all possible concatenations of substrokes. However, when the number of over-segmented substrokes increases, the merging speed will be greatly lowered because the search space of all possible combinations is large.…”

A Stroke Shape and Structure Based Approach for Off-line Chinese Handwriting Identification

“…Meanwhile there are more strokes in cursive style. According to [17], in Chinese characters, the complication structures are mostly affected by multi stokes of each character. Additionally, as shown in Figure 1, the stroke shapes and structures of Chinese characters are quite different from those of other languages such as English, which makes it more difficult to identify Chinese handwriting [6].…”

Recover Writing Trajectory from Multiple Stroked Image Using Bidirectional Dynamic Search

“…A template (model) is constructed to describe the shape of each stroke or the whole character(s) and the writing paths (strokes) are recovered by comparing the template with input image. Cheng [13] modeled the stroke by straight lines and extracted the strokes using relaxation labeling. Liu et al [2] classified stroke into different types according to length, orientation and relative position and used heuristic search to find the optimal paths.…”