• Journal of Information Processing Systems
• /
• v.18 no.3
• /
• pp.411-427
• /
• 2022

Digging historical and cultural information from seals in ancient books is of great significance. However, ancient Chinese seal samples are scarce and carving methods are diverse, and traditional digital image processing methods based on greyscale have difficulty achieving superior segmentation and recognition performance. Recently, some deep learning algorithms have been proposed to address this problem; however, current neural networks are difficult to train owing to the lack of datasets. To solve the afore-mentioned problems, we proposed an SEL-RefineMask which combines selector of feature pyramid network (SEL-FPN) with RefineMask to segment and recognize seals. We designed an SEL-FPN to intelligently select a specific layer which represents different scales in the FPN and reduces the number of anchor frames. We performed experiments on some instance segmentation networks as the baseline method, and the top-1 segmentation result of 64.93% is 5.73% higher than that of humans. The top-1 result of the SEL-RefineMask network reached 67.96% which surpassed the baseline results. After segmentation, a vision transformer was used to recognize the segmentation output, and the accuracy reached 91%. Furthermore, a dataset of seals in ancient Chinese books (SACB) for segmentation and small seal font (SSF) for recognition were established which are publicly available on the website.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.12
• /
• pp.3364-3382
• /
• 2023

Remote sensing image segmentation plays an important role in realizing intelligent city construction. The current mainstream segmentation networks effectively improve the segmentation effect of remote sensing images by deeply mining the rich texture and semantic features of images. But there are still some problems such as rough results of small target region segmentation and poor edge contour segmentation. To overcome these three challenges, we propose an improved semantic segmentation model, referred to as MRU-Net, which adopts the U-Net architecture as its backbone. Firstly, the convolutional layer is replaced by BasicBlock structure in U-Net network to extract features, then the activation function is replaced to reduce the computational load of model in the network. Secondly, a hybrid multi-scale recognition module is added in the encoder to improve the accuracy of image segmentation of small targets and edge parts. Finally, test on Massachusetts Buildings Dataset and WHU Dataset the experimental results show that compared with the original network the ACC, mIoU and F1 value are improved, and the imposed network shows good robustness and portability in different datasets.

• Proceedings of the IEEK Conference
• /
• 2000.11c
• /
• pp.145-148
• /
• 2000

This paper proposes a character segmentation method for Korean letter address image. The poor quality of image binarization results in broken character strokes. To overcome this problem, two steps of processing ate introduced. The first one is to merge broken characters to generate character candidates, and the other one is to reduce the complexity of segmentation graph path. These two steps do not use recognition information to keep in high-speed.

• Journal of the Semiconductor & Display Technology
• /
• v.20 no.3
• /
• pp.6-10
• /
• 2021

Recognition of characters printed or marked on the PCB components from images captured using cameras is an important task in PCB components inspection systems. Previous optical character recognition (OCR) of PCB components typically consists of two stages: character segmentation and classification of each segmented character. However, character segmentation often fails due to corrupted characters, low image contrast, etc. Thus, OCR without character segmentation is desirable and increasingly used via deep neural networks. Typical implementation based on deep neural nets without character segmentation includes convolutional neural network followed by recurrent neural network (RNN). However, one disadvantage of this approach is slow execution due to RNN layers. LPRNet is a segmentation-free character recognition network with excellent accuracy proved in license plate recognition. LPRNet uses a wide convolution instead of RNN, thus enabling fast inference. In this paper, LPRNet was adapted for recognizing characters printed on PCB components with fast execution and high accuracy. Initial training with synthetic images followed by fine-tuning on real text images yielded accurate recognition. This net can be further optimized on Intel CPU using OpenVINO tool kit. The optimized version of the network can be run in real-time faster than even GPU.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.8 no.7
• /
• pp.1480-1487
• /
• 2004

This study is proposed the novel gesture recognition method for the remote camera robot control. To recognize the dynamics gesture, the preprocessing step is the image segmentation. The conventional methods for the effectively object segmentation has need a lot of the cole. information about the object(hand) image. And these methods in the recognition step have need a lot of the features with the each object. To improve the problems of the conventional methods, this study proposed the novel method to recognize the dynamic hand gesture such as the MMS(Max-Min Search) method to segment the object image, MSM(Mean Space Mapping) method and COG(Conte. Of Gravity) method to extract the features of image, and the structure of recognition MLPNN(Multi Layer Perceptron Neural Network) to recognize the dynamic gestures. In the results of experiment, the recognition rate of the proposed method appeared more than 90[%], and this result is shown that is available by HCI(Human Computer Interface) device for .emote robot control.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.2
• /
• pp.13-20
• /
• 2024

This paper introduces a method for generating model images that can identify specific cylindrical medicine containers in videos and investigates data collection techniques. Previous research had separated object detection from specific object recognition, making it challenging to apply automated image stitching. A significant issue was that the coordinate-based object detection method included extraneous information from outside the object area during the image stitching process. To overcome these challenges, this study applies the newly released YOLOv8 (You Only Look Once) segmentation technique to vertically rotating pill bottles video and employs the ORB (Oriented FAST and Rotated BRIEF) feature matching algorithm to automate model image generation. The research findings demonstrate that applying segmentation techniques improves recognition accuracy when identifying specific pill bottles. The model images created with the feature matching algorithm could accurately identify the specific pill bottles.

• The Journal of the Korea Contents Association
• /
• v.18 no.4
• /
• pp.398-405
• /
• 2018

In the HCI, hand gesture recognition is attracting attention as a method for interaction and information exchange between users and devices along with the development of IT devices. In hand gesture recognition through image processing, palm region detection is a key process contributing to improvement of processing speed and recognition rate. In this paper, we propose a new method for image segmentation between the hand and wrist for palm area detection. The anatomical characteristics of the hand are used to calculate the distance between the iliac bones of the thumb and little finger, which have the widest width, by the horizontal projection histogram of the hand image, and then the palm area is detected by drawing a circle having the width as the diameter. In order to verify the superiority of this method, multiple stage template matching is used to compare and evaluate recognition performance against the four conventional methods for 10 hand gestures. Note that the literatures to offer palm area detection performance evaluation are few although there are many studies on hand gesture recognition.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.11
• /
• pp.5522-5540
• /
• 2018

Vehicular license plate (LP) recognition system has risen as a significant field of research recently because various explorations are currently being conducted by the researchers to cope with the challenges of LPs which include different illumination and angular situations. This research focused on restricted conditions such as using image of only one vehicle, stationary background, no angular adjustment of the skewed images. A real time vehicular LP recognition scheme is proposed for the skewed images for detection, segmentation and recognition of LP. In this research, a polar co-ordinate transformation procedure is implemented to adjust the skewed vehicular images. Besides that, window scanning procedure is utilized for the candidate localization that is based on the texture characteristics of the image. Then, connected component analysis (CCA) is implemented to the binary image for character segmentation where the pixels get connected in an eight-point neighbourhood process. Finally, optical character recognition is implemented for the recognition of the characters. For measuring the performance of this experiment, 300 skewed images of different illumination conditions with various tilt angles have been tested. The results show that proposed method able to achieve accuracy of 96.3% in localizing, 95.4% in segmenting and 94.2% in recognizing the LPs with an average localization time of 0.52s.

• Journal of Institute of Control, Robotics and Systems
• /
• v.6 no.1
• /
• pp.91-97
• /
• 2000

In this paper, the illuminator is designed to recognize the shape and the existence of holes of radiation fin in the point that the light reflection characteristics are different according to the roughness of the material. The threshold value, the positions of holes and the black pixel nembers in the positon are obtained under the illuminator, in accordance with the reference image, by applying binary conversion and hole segmentation algorithm, as they are suggested in this paper, The existence and shape of hole are recognized by calculating the distance and feature value in the test image, which is obtained from the parameters of reference image. It is programmed to apply to GUI(Graphic User the Interface) in windows. More than 98% of recognition rate is shown, as it is applied to three different sizes of the radiation fin.

• Proceedings of the IEEK Conference
• /
• 2003.11a
• /
• pp.469-472
• /
• 2003

Skew is inevitably occurred in a scanned document image Thus, character recognition systems are generally very sensitive to a skew angle. In this paper, we propose a robust slant correction algorithm based on dithering and estimating vortical transition. Character strings are segmented by projecting the vertical transition point and the slant corrected image. The segmentation method using the vertical transition point can effectively split the character strings touching vertically each other. Experimental results show that the proposed method has achieved robust slant correction and good performance of character string segmentation.