• Journal of information and communication convergence engineering
• /
• v.22 no.1
• /
• pp.80-87
• /
• 2024

The Manchu language holds historical significance, but a complete dataset of Manchu script letters for training optical character recognition machine-learning models is currently unavailable. Therefore, this paper describes the process of creating a robust dataset of extracted Manchu script letters. Rather than performing automatic letter segmentation based on whitespace or the thickness of the central word stem, an image of the Manchu script was manually inspected, and one copy of the desired letter was selected as a region of interest. This selected region of interest was used as a template to match all other occurrences of the same letter within the Manchu script image. Although the dataset in this study contained only 4,000 images of five Manchu script letters, these letters were collected from twenty-eight writing styles. A full dataset of Manchu letters is expected to be obtained through this process. The collected dataset was normalized and trained using a simple convolutional neural network to verify its effectiveness.

• Journal of Intelligence and Information Systems
• /
• v.24 no.4
• /
• pp.111-136
• /
• 2018

In this paper, we propose a methodology to extract answer information about queries from various types of unstructured documents collected from multi-sources existing on web in order to expand knowledge base. The proposed methodology is divided into the following steps. 1) Collect relevant documents from Wikipedia, Naver encyclopedia, and Naver news sources for "subject-predicate" separated queries and classify the proper documents. 2) Determine whether the sentence is suitable for extracting information and derive the confidence. 3) Based on the predicate feature, extract the information in the proper sentence and derive the overall confidence of the information extraction result. In order to evaluate the performance of the information extraction system, we selected 400 queries from the artificial intelligence speaker of SK-Telecom. Compared with the baseline model, it is confirmed that it shows higher performance index than the existing model. The contribution of this study is that we develop a sequence tagging model based on bi-directional LSTM-CRF using the predicate feature of the query, with this we developed a robust model that can maintain high recall performance even in various types of unstructured documents collected from multiple sources. The problem of information extraction for knowledge base extension should take into account heterogeneous characteristics of source-specific document types. The proposed methodology proved to extract information effectively from various types of unstructured documents compared to the baseline model. There is a limitation in previous research that the performance is poor when extracting information about the document type that is different from the training data. In addition, this study can prevent unnecessary information extraction attempts from the documents that do not include the answer information through the process for predicting the suitability of information extraction of documents and sentences before the information extraction step. It is meaningful that we provided a method that precision performance can be maintained even in actual web environment. The information extraction problem for the knowledge base expansion has the characteristic that it can not guarantee whether the document includes the correct answer because it is aimed at the unstructured document existing in the real web. When the question answering is performed on a real web, previous machine reading comprehension studies has a limitation that it shows a low level of precision because it frequently attempts to extract an answer even in a document in which there is no correct answer. The policy that predicts the suitability of document and sentence information extraction is meaningful in that it contributes to maintaining the performance of information extraction even in real web environment. The limitations of this study and future research directions are as follows. First, it is a problem related to data preprocessing. In this study, the unit of knowledge extraction is classified through the morphological analysis based on the open source Konlpy python package, and the information extraction result can be improperly performed because morphological analysis is not performed properly. To enhance the performance of information extraction results, it is necessary to develop an advanced morpheme analyzer. Second, it is a problem of entity ambiguity. The information extraction system of this study can not distinguish the same name that has different intention. If several people with the same name appear in the news, the system may not extract information about the intended query. In future research, it is necessary to take measures to identify the person with the same name. Third, it is a problem of evaluation query data. In this study, we selected 400 of user queries collected from SK Telecom 's interactive artificial intelligent speaker to evaluate the performance of the information extraction system. n this study, we developed evaluation data set using 800 documents (400 questions * 7 articles per question (1 Wikipedia, 3 Naver encyclopedia, 3 Naver news) by judging whether a correct answer is included or not. To ensure the external validity of the study, it is desirable to use more queries to determine the performance of the system. This is a costly activity that must be done manually. Future research needs to evaluate the system for more queries. It is also necessary to develop a Korean benchmark data set of information extraction system for queries from multi-source web documents to build an environment that can evaluate the results more objectively.

• Journal of KIISE:Software and Applications
• /
• v.30 no.10
• /
• pp.938-943
• /
• 2003

In this paper, we Propose a method using the Tikhonov-Miller process to improve the robustness of watermarking under various attacks. A visually recognizable pattern watermark is embedded in the LH2, HL2 and HH2 subband of wavelet transformed domain using threshold and besides watermark is embeded by utilizing HVS(Human Visual System) feature. The pattern watermark was interlaced after random Permutation for a security and an extraction rate. To demonstrate the improvement of robustness and similarity of the proposed method, we applied some basic algorithm of image processing such as scaling, filtering, cropping, histogram equalizing and lossy compression(JPEG, gif). As a result of experiment, the proposed method was able to embed robust watermark invisibility and extract with an excellent normalized correlation of watermark under various attacks.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.18 no.7
• /
• pp.935-943
• /
• 1993

In this paper, robust recognition system for tilted facial image was developed. At first, standard facial image and lilted facial image are captured by CCTV camera and then transformed into binary image. The binary image is processed in order to obtain contour image by Laplacian edge operator. We trace and delete outermost edge line and use inner contour lines. We label four inner contour lines in order among the inner lines, and then we extract left and right eye with known distance relationship and with two eyes coordinates, and calculate slope information. At last, we rotate the tilted image in accordance with slope information and then calculate the ten distance features between element and element. In order to make the system invariant to image scale, we normalize these features with distance between left and righ eye. Experimental results show 88% recognition rate for twenty five face images when tilted degree is considered and 60% recognition rate when tilted degree is not considered.

• Journal of Broadcast Engineering
• /
• v.17 no.2
• /
• pp.388-399
• /
• 2012

In the paper, we propose enhanced feature extraction and matching methods for a mobile environment based on modified SURF. We propose three methods to reduce the computational complexity in a mobile environment. The first is to reduce the dimensions of the SURF descriptor. We compare the performance of existing 64-dimensional SURF with several other dimensional SURFs. The second is to improve the performance using the sign of the trace of the Hessian matrix. In other words, feature points are considered as matched if they have the same sign for the trace of the Hessian matrix, otherwise considered not matched. The last one is to find the best distance-ratio which is used to determine the matching points. We find the best distance-ratio through experiments, and it gives the relatively high accuracy. Finally, existing system which is based on normal SURF method is compared with our proposed system which is based on these three proposed methods. We present that our proposed system shows reduced response time while preserving reasonably good matching accuracy.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.37 no.4
• /
• pp.11-19
• /
• 2000

Real-time extraction of traffic information such as the number of vehicles passing, speed, road-occupancy rate, distance between vehicles, and vehicle types from the traffic scenes acquired from the camera on the road, is a core component of the intelligent transportation system(lTS) We present a scheme of extracting the traffic informations based on the spatio-temporal image analysis, which is robust to the variation of weather conditions and the shades. The images of two detection regions for each traffic lane are classified into one of the three categories: the road, the vehicle, and the shade, using the statistical and structural features Quantitative traffic informations are retrieved by analysing the two spatio-temporal images. Since only the local images of detection regions are processed, the real-time operation of more than 30 frames per second is realized while ensuring the detection performance robust to the operating condition.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.4
• /
• pp.2124-2148
• /
• 2017

Accurate facial expression recognition (FER) requires reliable signal filtering and the effective feature extraction. Considering these requirements, this paper presents a novel approach for FER which is robust to noise. The main contributions of this work are: First, to preserve texture details in facial expression images and remove image noise, we improved the anisotropic diffusion filter by adjusting the diffusion coefficient according to two factors, namely, the gray value difference between the object and the background and the gradient magnitude of object. The improved filter can effectively distinguish facial muscle deformation and facial noise in face images. Second, to further improve robustness, we propose a new feature descriptor based on a combination of the Histogram of Oriented Gradients with the Canny operator (Canny-HOG) which can represent the precise deformation of eyes, eyebrows and lips for FER. Third, Canny-HOG's block and cell sizes are adjusted to reduce feature dimensionality and make the classifier less prone to overfitting. Our method was tested on images from the JAFFE and CK databases. Experimental results in L-O-Sam-O and L-O-Sub-O modes demonstrated the effectiveness of the proposed method. Meanwhile, the recognition rate of this method is not significantly affected in the presence of Gaussian noise and salt-and-pepper noise conditions.

• Journal of Internet Computing and Services
• /
• v.8 no.3
• /
• pp.33-43
• /
• 2007

In this proper, we proposes a robust watermarking against the lossy compression in the Hadamard domain. The Hadamard matrix consists of only 1 or -1 and can be computed veru fast. The Hadamrd transform has the inverse transform therefore it is able to be applied into the watermarking technology. In embedding process, we select 10 coefficients from intermediate frequency domain and create two watermark patterns. In extraction process, we use the watermark patterns and compare them to detect the watermark information. When we use the standard image ($512{\times}512$) and binary watermark image ($64{\times}64$), the results of these examines are PSNR for $38{\sim}42dB$ and BER for $3.9{\sim}12.5%$. The JPEG QF between 30 and100, naked human eyes can detect to watermark image easily. The experimental results show that performance of Hadamard domain is better than those of DCT, FFT, and DWT.

• Journal of Advanced Navigation Technology
• /
• v.16 no.2
• /
• pp.394-400
• /
• 2012

SURF is a robust local invariant feature descriptor and has been used in many applications such as object recognition. Even though this algorithm has similar matching accuracy compared to the SIFT, which is another popular feature extraction algorithm, it has advantage in matching time. However, these descriptors do not consider relative location information of extracted interesting points to guarantee rotation invariance. Also, since they use gray image of original color image, they do not use the color information of images, either. In this paper, we propose a method for improving matching performance of SURF descriptor using the color and relative location information of interest points. The location information is built from the angles between the line connecting the centers of interest points and the orientation line constructed for the center of each interest points. For the color information, color histogram is constructed for the region of each interest point. We show the performance of our scheme through experiments.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.41 no.12
• /
• pp.1978-1984
• /
• 2016

This paper proposes a robust vehicle detecting method by using Adaboost and CLAHE(Contrast-Limit Adaptive Histogram Equalization). We propose two method to detect vehicle effectively. First, we are able to judge rainy and night by converting RGB value to brightness. Second, we can detect a taillight, designate a ROI(Region Of Interest) by using CLAHE. And then, we choose an Adaboost algorithm by comparing traditional vehicle detecting method such as GMM(Gaussian Mixture Model), Optical flow and Adaboost. In this paper, we use proposed method and get better performance of detecting vehicle. The precision and recall score of proposed method are 0.85 and 0.87. That scores are better than GMM and optical flow.