• Journal of Korea Multimedia Society
• /
• v.21 no.12
• /
• pp.1467-1472
• /
• 2018

For the problem that the face image of surveillance video cannot be accurately identified due to the low resolution, this paper proposes a low resolution face recognition solution based on convolutional neural network model. Convolutional Neural Networks (CNN) model for multi-scale input The CNN model for multi-scale input is an improvement over the existing "two-step method" in which low-resolution images are up-sampled using a simple bi-cubic interpolation method. Then, the up sampled image and the high-resolution image are mixed as a model training sample. The CNN model learns the common feature space of the high- and low-resolution images, and then measures the feature similarity through the cosine distance. Finally, the recognition result is given. The experiments on the CMU PIE and Extended Yale B datasets show that the accuracy of the model is better than other comparison methods. Compared with the CMDA_BGE algorithm with the highest recognition rate, the accuracy rate is 2.5%~9.9%.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.12B
• /
• pp.1574-1585
• /
• 2011

This paper proposes a real-time algorithm for speed limit sign recognition for advanced safety vehicle system. The proposed algorithm uses Block-DCT in extracting features from a given ROI(Region Of Interest) instead of using entire pixel values as in previous works. The proposed algorithm chooses parts of the DCT coefficients according to the proposed discriminant factor, uses correlation coefficients and variances among ROIs from training samples to reduce amount of arithmetic operations without performance degradation in classification process. The algorithm recognizes the speed limit signs using the information obtained during training process by calculating LDA and Mahalanobis Distance. To increase the hit rate of recognition, it uses accumulated classification results computed for a sequence of frames. Experimental results show that the hit rate of recognition for sequential frames reaches up to 100 %. When compared with previous works, numbers of multiply and add operations are reduced by 69.3 % and 67.9 %, respectively. Start after striking space key 2 times.

• Journal of Platform Technology
• /
• v.11 no.1
• /
• pp.38-51
• /
• 2023

We propose an intelligent interface algorithm using hand gesture recognition information based on artificial intelligence. This method is functionally an interface that recognizes various motions quickly and intelligently by using MediaPipe and artificial intelligence techniques such as KNN, LSTM, and CNN to track and recognize user hand gestures. To evaluate the performance of the proposed algorithm, it is applied to a self-made 2D top-view racing game and robot control. As a result of applying the algorithm, it was possible to control various movements of the virtual object in the game in detail and robustly. And the result of applying the algorithm to the robot control in the real world, it was possible to control movement, stop, left turn, and right turn. In addition, by controlling the main character of the game and the robot in the real world at the same time, the optimized motion was implemented as an intelligent interface for controlling the coexistence space of virtual and real world. The proposed algorithm enables sophisticated control according to natural and intuitive characteristics using the body and fine movement recognition of fingers, and has the advantage of being skilled in a short period of time, so it can be used as basic data for developing intelligent user interfaces.

• International Journal of Aeronautical and Space Sciences
• /
• v.10 no.2
• /
• pp.86-94
• /
• 2009

This paper deals with the creation of a new, low-cost point/position tracking system that can measure deformations in engineering structures with simple commercially widespread cameras. Though point tracking systems do exist today, such as Stereo Pattern Recognition (SPR) and Projection Moir$\acute{e}$ Interferometry (PMI) systems, they are far too costly to use to analyze small, simple structures because complex optical components such as large flashes, high-resolution cameras and data acquisition systems with several computers are required. We developed a point tracking system using commercial cameras. This system used IR LEDs and commercial IR CCD cameras to minimize the interference posed by other extraneous light sources. The main algorithm used for this system is an optical point tracking algorithm, which is composed of the point extraction algorithm and the point matching algorithm for 3-D motion estimation. a series of verification tests were performed. Then, the developed point tracking system was applied to measure deformations of an acrylic plate under a mechanical load. The measured deformations of the acrylic plate matched well with the numerical analysis results. The results indicate that the developed point tracking system is reliable enough to measure continuous deformed shapes of various engineering structures.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.20 no.3
• /
• pp.368-374
• /
• 2010

Image segmentation is the basic step in the field of the image processing for pattern recognition, environment recognition, and context analysis. The Otsu's automatic threshold selection, which determines the optimal threshold value to maximize the between class scatter using the distribution information of the normalized histogram of a image, is the famous method among the various image segmentation methods. For the automatic threshold selection proposed by Otsu, it is difficult to determine the optimal threshold value by considering the sub-region characteristic of the image because the Otsu's algorithm analyzes the global histogram of a image. In this paper, to alleviate this difficulty of Otsu's image segmentation algorithm and to improve image segmentation capability, the original image is divided into several sub-images by using context fuzzy c-means algorithm. The proposed fuzzy Otsu threshold algorithm is applied to the divided sub-images and the several threshold values are obtained.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.40 no.6
• /
• pp.31-40
• /
• 2003

Most face recognition systems are based on 2D images and applied in many applications. However, it is difficult to recognize a face when the pose varies severely. Therefore, head pose estimation is an inevitable procedure to improve recognition rate when a face is not frontal. In this paper, we propose a novel head pose estimation algorithm for 3D face recognition. Given the 3D range image of an unknown face as an input, we automatically extract facial feature points based on the face curvature. We propose an Error Compensated Singular Value Decomposition (EC-SVD) method based on the extracted facial feature points. We obtain the initial rotation angle based on the SVD method, and perform a refinement procedure to compensate for remained errors. The proposed algorithm is performed by exploiting the extracted facial features in the normaized 3D face space. In addition, we propose a 3D nearest neighbor classifier in order to select face candidates for 3D face recognition. From simulation results, we proved the efficiency and validity of the proposed algorithm.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.42 no.4 s.304
• /
• pp.43-50
• /
• 2005

This paper proposes a new pattern recognition system combining the new adaptive feature weighting based on the genetic algorithm and the modified KNN(K Nearest-Neighbor) rules. The new feature weighting proposed herein avoids the overfitting and finds the Proper feature weighting value by determining the middle value of weights using GA. New GA operators are introduced to obtain the high performance of the system. Moreover, a class dependent feature weighting strategy is employed. Whilst the classical methods use the same feature space for all classes, the Proposed method uses a different feature space for each class. The KNN rule is modified to estimate the class of test pattern using adaptive feature space. Experiments were performed with the unconstrained handwritten numeral database of Concordia University in Canada to show the performance of the proposed method.

• Journal of Korea Multimedia Society
• /
• v.7 no.8
• /
• pp.1035-1045
• /
• 2004

In this paper, we describe a method that can recognize gestures and evaluate the degree of the gestures from sequential gesture images by using Gesture Feature Space. The previous popular methods based on HMM and neural network have difficulties in recognizing the degree of gesture even though it can classify gesture into some kinds. However, our proposed method can recognize not only posture but also the degree information of the gestures, such as speed and magnitude by calculating distance among the position vectors substituting input and model images in parametric eigenspace. This method which can be applied in various applications such as intelligent interface systems and surveillance systems is a simple and robust recognition algorithm.

• The Journal of the Korea Contents Association
• /
• v.5 no.2
• /
• pp.267-274
• /
• 2005

In this paper, we proposed an algorithm which cognize hand pose in real time using only image. Hand recognizes using edge orientation histogram which comes under a constant quantity of 2D appearance because hand pose is intricate. This method suit hand pose recognition in real time because it extracts hand space accurately, has little computation quantify, and is less sensitive to lighting change using color information in complicated background. Method which reduces recognition error using principal component analysis method to can recognize through hand shape presentation direction change is explained. A case that hand shape changes by turning 3D also by using this method is possible to recognize. Besides, principal component space creation time is reduced remarkably because edge directional data is used.

• Journal of Institute of Control, Robotics and Systems
• /
• v.16 no.2
• /
• pp.158-163
• /
• 2010

In this paper, we propose the location measurement algorithm of unknown tag based on RFID (Radio-Frequency IDentification) by using RSSI (Received Signal Strength Indication) and TDOA (Time Difference of Arrival) and extended Kalman filter in smart space. To do this, first, we recognize the location of unknown tag by using the RSSI and TDOA recognition methods. Second, we set the coordinate of the tag location measured by using trilateration and SX algorithm. But the tag location data measured by this method are included complex environmental error. So, we use the extended Kalman filter in order to revise error data of the tag location. Finally, we validate the applicability of the proposed method though the simulation in a complex environment.