• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.2
• /
• pp.78-88
• /
• 2009

In many face recognition problems, the number of available images is limited compared to the dimension of the input space which is usually equal to the number of pixels. This problem is called as the 'small sample size' problem and regularization methods are typically used to solve this problem in feature extraction methods such as LDA. By using regularization methods, the modified within class matrix becomes nonsingu1ar and LDA can be performed in its original form. However, in the process of adding a scaled version of the identity matrix to the original within scatter matrix, the scale factor should be set heuristically and the performance of the recognition system depends on highly the value of the scalar factor. By using the proposed resampling method, we can generate a set of images similar to but slightly different from the original image. With the increased number of images, the small sample size problem is alleviated and the classification performance increases. Unlike regularization method, the resampling method does not suffer from the heuristic setting of the parameter producing better performance.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.26 no.10
• /
• pp.1445-1450
• /
• 2002

In the present study, a new pattern recognition algorithm was proposed to size spray particles using the boundary curvature information. Conceptually, this algorithm has an advantage over the others because it can identify the particle size and shape simultaneously, and also can separate the overlapped particles more effectively. Curvature of a boundary was obtained from the change of the slopes of two neighboring segments at the corresponding part. The algorithm developed in this study was tested by using an artificially prepared image of a group of spherical particles which were either isolated or overlapped. Particle sizes obtained from the measured curvatures agreed well with the true values. By detecting abrupt changes of the curvature along the image boundary, the element particles could be separated out from their overlapped images successfully.

• Convergence Security Journal
• /
• v.19 no.1
• /
• pp.67-78
• /
• 2019

Most existing facial expression recognition methods use a uniform grid method that divides the entire facial image into uniform blocks when describing facial features. The problem of this method may include non-face backgrounds, which interferes with discrimination of facial expressions, and the feature of a face included in each block may vary depending on the position, size, and orientation of the face in the input image. In this paper, we propose a variable-size block method which determines the size and position of a block that best represents meaningful facial expression change. As a part of the effort, we propose the way to determine the optimal number, position and size of each block based on the facial feature points. For the evaluation of the proposed method, we generate the facial feature vectors using LDTP and construct a facial expression recognition system based on SVM. Experimental results show that the proposed method is superior to conventional uniform grid based method. Especially, it shows that the proposed method can adapt to the change of the input environment more effectively by showing relatively better performance than exiting methods in the images with large shape and orientation changes.

• Journal of Intelligence and Information Systems
• /
• v.9 no.2
• /
• pp.1-18
• /
• 2003

The automatic recognition of transport containers using image processing is very hard because of the irregular size and position of identifiers, diverse colors of background and identifiers, and the impaired shapes of identifiers caused by container damages and the bent surface of container, etc. We proposed and evaluated the novel recognition algorithm of container identifiers that overcomes effectively the hardness and recognizes identifiers from container images captured in the various environments. The proposed algorithm, first, extracts the area including only all identifiers from container images by using CANNY masking and bi-directional histogram method. The extracted identifier area is binarized by the fuzzy binarization method newly proposed in this paper and by applying contour tracking method to the binarized area, container identifiers which are targets of recognition are extracted. We proposed and applied the ART2-based RBF network for recognition of container identifiers. The results of experiment for performance evaluation on the real container images showed that the proposed algorithm has more improved performance in the extraction and recognition of container identifiers than the previous algorithms.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.32B no.8
• /
• pp.1110-1118
• /
• 1995

Over the past few years, a wide variety of techniques have been developed which make a reliable recognition of speech signal. Multi-layer perceptron(MLP) which has excellent pattern recognition properties is one of the most versatile networks in the area of speech recognition. This paper describes an automatic speech recognition system which use both MLP and lexicon. In this system., the recognition is performed by a network search algorithm which matches words in lexicon to MLP output scores. We also suggest a recognition algorithm which incorperat durational information of each phone, whose performance is comparable to that of conventional continuous HMM(CHMM). Performance of the system is evaluated on the database of 26 vocabulary size from 9 speakers. The experimental results show that the proposed algorithm achieves error rate of 7.3% which is 5.3% lower rate than 12.6% of CHMM.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.35C no.8
• /
• pp.60-67
• /
• 1998

We studied the effects of fractional fourier transform in face recognition, in which only the amplitude spectra of transformed facial images were used.We used two recently developed face recognition methods, the most effective feature (MEF) method (i.e., eigenface method) and most discriminating feature (MDF) method, and the effects of th etransform for th etwo methods were consistent. We confirmed that the recognition rate by the use of MDF method is better than that consistent. We confirmed that the recognition rate by the use of MDF method is better than that by MEF regardless of the order to transform, these methods provided slightly better results when the order was 1 than for any other order values. Only when the order was close to 1, the recognition rates were robust to the shift of the input images, and the trend that the recognition rates decreased as the input size varied was independent of the order. From these results, we fond that it is most advantageous to use the amplitude spectra of the conventional fourier transform whose order is 1.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.57-60
• /
• 2001

This paper reports on speaker-Recognition of context independence-speaker recognition in the field of the speech recognition. It is important to select the parameter reflecting the characteristic of each single person because speaker-recognition is to identify who speaks in the database. We used Mel Frequency Cesptrum Coefficient and Vector Quantization to identify in this paper. Specially, it considered to find characteristic-vector of the speaker in different from known method; this paper used the characteristic-vector which is selected in MFCC Parameter Space. Also, this paper compared the recognition rate according to size of codebook from this database and the time needed for operation with the existing one. The results is more improved $3\sim4\%$ for recognition rate than established Vector Quantization Algorithm.

• Journal of information and communication convergence engineering
• /
• v.2 no.3
• /
• pp.177-181
• /
• 2004

In this paper, we suggest an Iris recognition system with an excellent recognition rate and confidence as an alternative biometric recognition technique that solves the limit in an existing individual discrimination. For its implementation, we extracted coefficients feature values with the wavelet transformation mainly used in the signal processing, and we used neural network to see a recognition rate. However, Scale Conjugate Gradient of nonlinear optimum method mainly used in neural network is not suitable to solve the optimum problem for its slow velocity of convergence. So we intended to enhance the recognition rate by using Levenberg-Marquardt Back-propagation which supplements existing Scale Conjugate Gradient for an implementation of the iris recognition system. We improved convergence velocity, efficiency, and stability by changing properly the size according to both convergence rate of solution and variation rate of variable vector with the implementation of an applied algorithm.

• Journal of Korea Multimedia Society
• /
• v.25 no.6
• /
• pp.807-818
• /
• 2022

In order to further improve the accuracy and time efficiency of behavior recognition in intelligent monitoring scenarios, a human behavior recognition algorithm based on YOLO combined with LSTM and CNN is proposed. Using the real-time nature of YOLO target detection, firstly, the specific behavior in the surveillance video is detected in real time, and the depth feature extraction is performed after obtaining the target size, location and other information; Then, remove noise data from irrelevant areas in the image; Finally, combined with LSTM modeling and processing time series, the final behavior discrimination is made for the behavior action sequence in the surveillance video. Experiments in the MSR and KTH datasets show that the average recognition rate of each behavior reaches 98.42% and 96.6%, and the average recognition speed reaches 210ms and 220ms. The method in this paper has a good effect on the intelligence behavior recognition.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.1
• /
• pp.17-23
• /
• 2005

In this paper, we propose face recognition method using embedded data in super states segmentalized that is specification region exist to face region, hair, forehead, eyes, ears, nose, mouth, and chin. Proposed method defines super states that is specification area in normalized size (92×112), and embedded data that is extract internal factor in super states segmentalized achieve face recognition by PCA algorithm. Proposed method can receive specification data that is less in proposed image's size (92×112) because do orignal image to learn embedded data not to do all loaming. And Showed face recognition rate in image of 92×112 size averagely 99.05%, step 1 99.05%, step 2 98.93%, step 3 98.54%, step 4 97.85%. Therefore, method that is proposed through an experiment showed that the processing speed improves as well as reduce existing face image's information.