• Journal of the Korea Society of Computer and Information
• /
• v.25 no.1
• /
• pp.55-61
• /
• 2020

With the development of deep learning technology and advances in computing power, video-based research is now gaining more and more attention. Video data contains a large amount of temporal and spatial information, which is the biggest difference compared with image data. It has a larger amount of data. It has attracted intense attention in computer vision. Among them, motion recognition is one of the research focuses. However, the action recognition of human in the video is extremely complex and challenging subject. Based on many research in human beings, we have found that artificial intelligence-like attention mechanisms are an efficient model for cognition. This efficient model is ideal for processing image information and complex continuous video information. We introduce this attention mechanism into video action recognition, paying attention to human actions in video and effectively improving recognition efficiency. In this paper, we propose a new 3D residual attention network using convolutional neural network based on two attention models to identify human action behavior in the video. An evaluation result of our model showed up to 90.7% accuracy.

• Proceedings of the Korea Contents Association Conference
• /
• 2009.05a
• /
• pp.105-111
• /
• 2009

Widely used color information recognition methods based on the RGB color model with static fuzzy inference rules have limitations due to the model itself - the detachment of human vision and applicability of limited environment. In this paper, we propose a method that is based on HSI model with new inference process that resembles human vision recognition process. Also, a user can add, delete, update the inference rules in this system. In our method, we design membership intervals with sine, cosine function in H channel and with functions in trigonometric style in S and I channel. The membership degree is computed via interval merging process. Then, the inference rules are applied to the result in order to infer the color information. Our method is proven to be more intuitive and efficient compared with RGB model in experiment.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.12
• /
• pp.6069-6091
• /
• 2017

Augmented Reality (AR) overlays virtual information on real world data, such as displaying useful information on videos/images of a scene. This paper presents an Enhanced AR (EAR) system that displays useful statistical players' information on captured images of a sports game. We focus on the situation where the input image is degraded by strong sunlight. Proposed EAR system consists of an image enhancement technique to improve the accuracy of subsequent player and face detection. The image enhancement is followed by player and face detection, face recognition, and players' statistics display. First, an algorithm based on multi-scale retinex is proposed for image enhancement. Then, to detect players' and faces', we use adaptive boosting and Haar features for feature extraction and classification. The player face recognition algorithm uses boosted linear discriminant analysis to select features and nearest neighbor classifier for classification. The system can be adjusted to work in different types of sports where the input is an image and the desired output is display of information nearby the recognized players. Simulations are carried out on 2096 different images that contain players in diverse conditions. Proposed EAR system demonstrates the great potential of computer vision based approaches to develop AR applications.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.10
• /
• pp.1280-1287
• /
• 2020

This paper proposes a vision-based food recognition method for automated production of custom cakes. A small camera module mounted on a food art printer recognizes objects' shape and estimates their center points through image processing. Through the perspective transformation, the top-view image is obtained from the original image taken at an oblique position. The line and circular hough transformations are applied to recognize square and circular shapes respectively. In addition, the center of gravity of each figure are accurately detected in units of pixels. The test results show that the shape recognition rate is more than 98.75% under 180 ~ 250 lux of light and the positioning error rate is less than 0.87% under 50 ~ 120 lux. These values sufficiently meet the needs of the corresponding market. In addition, the processing delay is also less than 0.5 seconds per frame, so the proposed algorithm is suitable for commercial purpose.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.4
• /
• pp.47-57
• /
• 2009

Event detection and recognition is an active and challenging topic recent in Computer Vision. This paper describes a new method for recognizing events caused by human motion from video sequences in an office environment. The proposed approach analyzes human motions using Motion History Image (MHI) sequences, and is invariant to body shapes. types or colors of clothes and positions of target objects. The proposed method has two advantages; one is thant the proposed method is less sensitive to illumination changes comparing with the method using color information of objects of interest, and the other is scale invariance comparing with the method using a prior knowledge like appearances or shapes of objects of interest. Combined with edge detection, geometrical characteristics of the human shape in the MHI sequences are considered as the features. An advantage of the proposed method is that the event detection framework is easy to extend by inserting the descriptions of events. In addition, the proposed method is the core technology for event detection systems based on context-aware computing as well as surveillance systems based on computer vision techniques.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.8
• /
• pp.25-31
• /
• 2009

Widely used color information recognition methods based on the RGB color model with static fuzzy inference rules have limitations due to the model itself-the detachment of human vision and applicability of limited environment. In this paper, we propose a method that is based on HSI model with new inference process that resembles human vision recognition process. Also, a user can add, delete, update the inference rules in this system. In our method, we design membership intervals with sine, cosine function in H channel and with functions in trigonometric style in S and I channel. The membership degree is computed via interval merging process. Then, the inference rules are applied to the result in order to infer the color information. Our method is proven to be more intuitive and efficient compared with RGB model in experiment.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.1
• /
• pp.86-99
• /
• 2011

Due to the uncertainty of intention recognition for behaviors of users, the intention is differently recognized according to the situation for the same behavior by the same user, the accuracy of user intention recognition by minimizing the uncertainty is able to be improved. This paper suggests a novel ontology-based method to recognize user intentions, and able to minimize the uncertainties that are the obstacles against the precise recognition of user intention. This approach creates ontology for user intention, makes a hierarchy and relationship among user intentions by using RuleML as well as Dynamic Bayesian Network, and improves the accuracy of user intention recognition by using the defined RuleML as well as the gathered sensor data such as temperature, humidity, vision, and auditory. To evaluate the performance of robot proactive planning mechanism, we developed a simulator, carried out some experiments to measure the accuracy of user intention recognition for all possible situations, and analyzed and detailed described the results. The result of our experiments represented relatively high level the accuracy of user intention recognition. On the other hand, the result of experiments tells us the fact that the actions including the uncertainty get in the way the precise user intention recognition.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.1413-1415
• /
• 2004

This paper proposes a SVM(Support Vector Machine) training algorithm to control a service robot with voice command. The service robot with a stereo vision system and dual manipulators of four degrees of freedom implements a User-Dependent Voice Control System. The training of SVM algorithm that is one of the statistical learning theories leads to a QP(quadratic programming) problem. In this paper, we present an efficient SVM speech recognition scheme especially based on less learning data comparing with conventional approaches. SVM discriminator decides rejection or acceptance of user's extracted voice features by the MFCC(Mel Frequency Cepstrum Coefficient). Among several SVM kernels, the exponential RBF function gives the best classification and the accurate user recognition. The numerical simulation and the experiment verified the usefulness of the proposed algorithm.

• Proceedings of the Korean Institute of Communication Sciences Conference
• /
• 1986.10a
• /
• pp.45-52
• /
• 1986

This paper presents an experimental model-based vision system which can identify and locate object in scenes containing multiple occluded parts. The objent are assumed to be regid, planar parta. In any recognition system the type of object that might appear in the image dictates the type of knowledge that is needed to recognize the object. The data is reduced to a seguential list of points or pixel that appear on the boundary of the objects. Next the boundary of the object is smoothed using a polygonal approximation algorithm. Recognition consists in finding the prototype that matches model to image. The best match is obtained by optimising some similarity measure.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.105.2-105
• /
• 2001

This paper studies continuous Korean Sign Language (KSL) recognition using color vision. In recognizing gesture words such as sign language, it is a very difficult to segment a continuous sign into individual sign words since the patterns are very complicated and diverse. To solve this problem, we disassemble the KSL into 18 hand motion classes according to their patterns and represent the sign words as some combination of hand motions. Observing the speed and the change of speed of hand motion and using state automata, we reject unintentional gesture motions such as preparatory motion and meaningless movement between sign words. To recognize 18 hand motion classes we adopt Hidden Markov Model (HMM). Using these methods, we recognize 5 KSL sentences and obtain 94% recognition ratio.