Makoto Yasuhara

Recovery of drawing order from single-stroke handwriting images

Describes a method to recover a drawing order of a handwritten script from a static 2D image. The script should be written in a single stroke and may include double-traced lines. After the script is scanned in and preprocessed, we apply our recovery method which consists of two phases. In the first phase, we globally analyze the graph constructed from the skeletal image and label the graph by determining the types of each edge. In the second phase, we trace the graph from the start vertex to the end vertex using the labeling information. This method does not enumerate the possible cases, for example, by solving the traveling salesman problem and, therefore, does not cause a combinatorial explosion even if the script is very complex. By recovering a drawing order of a handwritten script, the temporal information can be recovered from a static 2D image. Hence, this method will be used as a bridge from the offline handwriting character recognition problem to the online one.

A Framework Toward Restoration of Writing Order from Single-Stroked Handwriting Image

Restoration of writing order from a single-stroked handwriting image can be seen as the problem of finding the smoothest path in its graph representation. In this paper, a 3-phase approach to restore a writing order is proposed within the framework of the edge continuity relation (ECR). In the initial, local phase, in order to obtain possible ECRs at an even-degree node, a neural network is used for the node of degree 4 and a theoretical approach is presented for the node of degree higher than 4 by introducing certain reasonable assumptions. In the second phase, we identify double-traced lines by employing maximum weighted matching. This makes it possible to transform the problem of obtaining possible ECRs at odd-degree node to that at even-degree node. In the final, global phase, we find all the candidates of single-stroked paths by depth first search and select the best one by evaluating SLALOM smoothness. Experiments on static images converted from online data in the Unipen database show that our method achieves a restoration rate of 96.0 percent

Recovery of drawing order from scanned images of multi-stroke handwriting

Describes a method to recover the drawing order of a multi-stroke handwritten script from a binary 2D image. First, we construct a graph from the scanned image by applying a thinning process and extracting the skeletal pixels. Next, we identify the start or end vertices in the graph, and then globally analyze the graph to label it by determining the types of each vertex and each edge. Finally, we trace all the strokes using the labeling information and recover the drawing order. The method does not enumerate the possible drawing orders and does not cause a combinatorial explosion, even if the script is very complex. By recovering the drawing order of a handwritten script, the temporal information can be recovered from a scanned image. Hence, this method can be used as a bridge from the offline handwriting character recognition problem to the online one.

Recover Writing Trajectory from Multiple Stroked Image Using Bidirectional Dynamic Search

The recovery of writing trajectory from offline handwritten image is generally regarded as a difficult problem (Plamondon and Srihari, 2000). This paper introduced a method to recover the writing trajectory from multiple stroked images by searching the best matching writing paths of template strokes. The searching procedure is guided by a matching cost function which is defined as the summation of positional distortion cost and directional difference cost between the template stroke and its matching path. We develop a bidirectional search algorithm based on dynamic programming to find the best matching path. The algorithm can efficiently reduce the searching space, while hold the start/end vertex constraint. Experiments on the handwritten English words and Chinese characters demonstrated the effectiveness of our method

Cursive Character Recognition System

During the last two decade, numerous handwriting character recognition systems have been proposed. Many of them presented their limitation when the handwriting character is cursive type and it has some deformation. However this type of cursive character is easily recognized by the human being. In this paper we research its human ability and apply it to the dynamic handwriting character recognition. In the proposed system, significant knots of each character are extracted using natural spline function named SLALOM and their position is optimized steepest descent method. Using a training set consisting of the sequence of optimal knots, each character model is constructed. Finally the unknown input character is compared with each model of all characters to get the similarity scores. The character model with higher similarity score is considered as the recognized character of the input data. The recognition stage consists in two-steps: classification using global feature and classification using local feature

An Improved Adaptive Predictor in DPCM Based on the Kalman Filter and Its Application to Handwriting Signal Encoding

A method of handwriting signal encoding based on adaptive linear predictive coding (ALPC) is studied. The ALPC is a form of DPCM which uses a sequentially adaptive predictor in which a sequential estimation algorithm is used to update predictor coefficients. To improve the estimates of the predictor coefficients in the presence of quantization noise, Kalman filtering has been investigated for its feasibility. This results in improvements of not only the estimation of the predictor coefficients, but the signal-to-quantization-noise ratio (SNR) of the signals reconstructed at the receiver as well. Computer simulations have verified that the ALPC system employing the Kalman filter promises high performance and feasibility at the rate of 192 bits/s when applied to handwriting signal encoding.

Optimal Euler Circuit of Maximum Contiguous Cost

This paper introduces a new graph problem to find an Optimal Euler Circuit (OEC) in an Euler graph. OEC is defined as the Euler circuit that maximizes the sum of contiguous costs along it, where the contiguous cost is assigned for each of the two contiguous edges incident to a vertex. We prove that the OEC problem is NP-complete. A polynomial time algorithm will be presented for the case of a graph without vertex of degree greater than 4, and for the general case, a 1/4-approximation polynomial time algorithm will be proposed.

Novel Cursive Character Recognition System

During the last two decade, numerous handwriting character recognition systems have been proposed. Many of them presented their limitation when the handwriting character is cursive type and it has some deformation. However this type of cursive character is easily recognized by the human being. In this paper we research its human ability and apply it to the dynamic handwriting character recognition. In the proposed system, significant knots of each character are extracted using natural Spline function named SLALOM and their position is optimized with Steepest Descent Method. Using a training set consisting of the sequence of optimal knots, each character model will be constructed. Finally the unknown input character will be compared with each model of all characters to get the similarity scores. The character model with higher similarity score will be considered as the recognized character of the input data. The recognition stage consists in two-steps: classification using global feature and classification using local feature. The global recognition rate of the proposed system is approximately 96%.

A system for on-line handwritten character recognition using natural spline function

During the last several years there have been developed many systems which are able to simulate the human brain behavior. To achieve this goal, two of the most important paradigms used, are the Neural Networks and the Artificial Intelligence. Both of them are primary tools for development of systems to capable of performing tasks such as: handwritten characters, voice, faces, signatures recognition and so many other biometric applications that have attracted considerable attention during the last few years. In this paper a new algorithm for cursive handwritten characters recognition based on the Spline function is proposed, in which the inverse order of the handwritten character construction task will be used to recognize the character. From the sampled data obtained by using a digitizer board, the sequence of the most significant points (optimal knots) of the handwriting character will be obtain, and then the natural Spline function and the steepest descent method will be used to interpolate and approximate character shape, Using a training set consisting of the sequence of optimal knots, each character model will be constructed. Finally the unknown input character will be compared by all characters models to get the similitude scores. The character model with higher similitude score will be considered as the recognized character of the input data. The proposed system is evaluated by computer simulation and simulation results show the global recognition rate with 93.5%.