• Journal of Information Processing Systems
• /
• v.17 no.1
• /
• pp.1-13
• /
• 2021

In this research, we study the problem of font image skeletonization using an end-to-end deep adversarial network, in contrast with the state-of-the-art methods that use mathematical algorithms. Several studies have been concerned with skeletonization, but a few have utilized deep learning. Further, no study has considered generative models based on deep neural networks for font character skeletonization, which are more delicate than natural objects. In this work, we take a step closer to producing realistic synthesized skeletons of font characters. We consider using an end-to-end deep adversarial network, SkelGAN, for font-image skeletonization, in contrast with the state-of-the-art methods that use mathematical algorithms. The proposed skeleton generator is proved superior to all well-known mathematical skeletonization methods in terms of character structure, including delicate strokes, serifs, and even special styles. Experimental results also demonstrate the dominance of our method against the state-of-the-art supervised image-to-image translation method in font character skeletonization task.

• Journal of Computing Science and Engineering
• /
• v.8 no.3
• /
• pp.149-156
• /
• 2014

Skeletonization is a morphological operation that transforms an original object into a subset, which is called a 'skeleton'. Skeletonization has been intensively studied for decades and is a challenging issue especially for special target objects. This paper proposes a novel approach to the skeletonization of text images based on stroke width detection. First, the preliminary skeleton is detected by using a Canny edge detector with a Tensor Voting framework. Second, the preliminary skeleton is smoothed, and junction points are connected by interpolation compensation. Experimental results show the validity of the proposed approach.

• Journal of applied mathematics & informatics
• /
• v.15 no.1_2
• /
• pp.425-434
• /
• 2004

Pattern recognition in three dimensional image is highly sensitive to assigned value and formation of voxels (pixels for two dimension case). However, occurred while digital imaging, digitization error leads to unpredictable noises in image data. Skeletonization, a powerful tool of pattern recognition, is sensitively dependent on boundary formation. Without successful controlling of the noises, the results of skeletonization can not be allowed as a stable solution. To minimize the effect of noises affecting to boundary formation, we developed a robust processing method useful in skeletonization technique for pattern recognition. Finally, we provide rigorous test results achieved throughout simulation on analytic three dimensional image.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.1
• /
• pp.358-377
• /
• 2015

In this paper, we propose a novel approach for trajectory recovery. Our system uses a triangulation procedure for skeletonization and graph theory to extract the trajectory. Skeletonization extracts the polyline skeleton according to the polygonal contours of the handwritten characters, and as a result, the junction becomes clear and the characters that are touching each other are separated. The approach for the trajectory recovery is based on graph theory to find the optimal path in the graph that has the best representation of the trajectory. An undirected graph model consisting of one or more strokes is constructed from a polyline skeleton. By using the polyline skeleton, our approach accelerates the process to search for an optimal path. In order to evaluate the performance, we built our own dataset, which includes testing and ground-truth. The dataset consist of thousands of handwritten characters and word images, which are extracted from five handwritten documents. To show the relative advantage of our skeletonization method, we first compare the results against those from Zhang-Suen, a state-of-the-art skeletonization method. For the trajectory recovery, we conduct a comparison using the Root Means Square Error (RMSE) and Dynamic Time Warping (DTW) in order to measure the error between the ground truth and the real output. The comparison reveals that our approach has better performance for both the skeletonization stage and the trajectory recovery stage. Moreover, the processing time comparison proves that our system is faster than the existing systems.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2004.05b
• /
• pp.521-526
• /
• 2004

It is necessary to analyze the unsteady flow in the pipe network for the better operation and controls, but there are some problems in actual pipe network simulation, such as collecting a large amount of information in the field, operating highly upgraded computer system, and keeping a big storage device to run analysis program. The skeletonization method is used to cope with the problems in this paper. It is expected to reduce computation time, researcher's efforts, and costs for the analyzing the pipe network. The impact of individual pipe elements to the behavior of the water distribution system can be accounted in the process of skeletonization. However it is also important to study continuously about how to apply the skeletonization method for each of different cases, because inadequate uses may bring simulation to a false result. This paper introduces basic theories and skeletonizing examples in the actual pipe network in Dae-gu city.

• Journal of Korea Water Resources Association
• /
• v.48 no.10
• /
• pp.845-855
• /
• 2015

Studies of optimizing pump operation in water distribution networks (WDN) are receiving spotlight in recent days. However, the water networks are quite complex including thousands of or even tens of thousands of nodes and pipes, thus simulation time is an issue. In some cases, implementing a computer model for pump operation decisions is restrictive due to intensive computation time. To that end, it is necessary to reduce the simulation time of water networks by simplifying the network layout. In this study, WDN skeletonization approaches were suggested and applied to a real water transmission network in South Korea. In skeletonizing the original network, it was constrained to match the water pressure and water age in the same junction locations to maintain the hydraulic and water quality characteristics in the skeletonized network. Using the skeletonization approaches suggested in this study, it is expected to reduce the simulation time of WDN and apply for developing a computer module of WDN real-time optimal operation.

• Journal of Chest Surgery
• /
• v.28 no.7
• /
• pp.671-677
• /
• 1995

Skeletonization of the internal mammary artery [IMA during myocardial revascularization procedures may provide some advantages, compared with the pedicle graft of the artery. In 17 patients undergoing IMA grafting by skeletonization technique, flow through the artery was measured on mean arterial pressure of 50-55 mmHg immediately after cardiopulmonary bypass started [first flow and just before its anastomosis to left anterior descending artery [second flow . In 16 patients except 1 patient whose graft was injured during mobilization, the first flow of IMA graft was 32.3 $\pm$ 7.4 ml/min and the second flow increased to 59.6$\pm$25.9 ml/min without any treatment and the site for anastomosis of the IMA graft was more than 1.0 cm above the bifurcation. On the basis of previous clinical studies, the flow of the skeletonized IMA was greater than that of the pedicle graft [59.6 $\pm$ 25.9 ml/min versus 37.7$\pm$ 14.1 ml/min, p < 0.05 . In comparison between the skeletonized IMA and the IMA graft intraluminally dilated with papaverine solution, there was no significant difference between two flows[59.6 $\pm$25.7 ml/min versus 74.7 $\pm$31.4 ml/min, not significant , but the former showed longer graft and anastomosis of more proximal portion of the graft to left anterior descending artery. In conclusion, the technique of internal mammary artery skeletonization has consistently produced a satifactory conduit for myocardial revascularization procedures. We have adopted IMA skeletonization not only because of the flow, diameter, and vessel length obtained but also because of limited perivascular tissue disruption that occurs during the dissection.

• Proceedings of the KIEE Conference
• /
• 2004.11c
• /
• pp.334-336
• /
• 2004

The postprocessing of fingerprint images is widely used in the elimination of the false minutiae caused by skeletonization. This paper presents the images were duplicated by The SOFM. And this Method showed that the good performance of eliminating false minutiae and fast processing.

• The Journal of Korea Robotics Society
• /
• v.13 no.2
• /
• pp.142-149
• /
• 2018

$RRT^*$ (Rapidly exploring Random $Tree^*$) based algorithms are widely used for path planning. Informed $RRT^*$ uses $RRT^*$ for generating an initial path and optimizes the path by limiting sampling regions to the area around the initial path. $RRT^*$ algorithms have several limitations such as slow convergence speed, large memory requirements, and difficulties in finding paths when narrow aisles or doors exist. In this paper, we propose an algorithm to deal with these problems. The proposed algorithm applies the image skeletonization to the gridmap image for generating an initial path. Because this initial path is close to the optimal cost path even in the complex environments, the cost can converge to the optimum more quickly in the proposed algorithm than in the conventional Informed $RRT^*$. Also, we can reduce the number of nodes and memory requirement. The performance of the proposed algorithm is verified by comparison with the conventional Informed $RRT^*$ and Informed $RRT^*$ using initial path generated by $A^*$.

• Bulletin of the Korean Chemical Society
• /
• v.22 no.12
• /
• pp.1291-1292
• /
• 2001