• Journal of Korea Multimedia Society
• /
• v.12 no.9
• /
• pp.1252-1261
• /
• 2009

One notable challenge in video searching and summarizing is extracting semantic from video contents and annotating context for video contents. Video semantic or context could be obtained by two methods to extract objects and contexts between objects from video. However, the method that use just to extracts objects do not express enough semantic for shot or scene as it does not describe relation and interaction between objects. To be more effective, after extracting some objects, context like relation and interaction between objects needs to be extracted from conversation situation. This paper is a study for how to detect speaker and how to compose context for talking to annotate conversation context. For this, based on this study, we proposed the methods that characters are recognized through face recognition technology, speaker is detected through mouth motion, conversation context is extracted using the rule that is composed of speaker existing, the number of characters and subtitles existing and, finally, scene context is changed to xml file and saved.

• Journal of the Institute of Convergence Signal Processing
• /
• v.2 no.2
• /
• pp.28-38
• /
• 2001

Traffic information can be broadly categorized into point information and spatial information. Point information can be obtained by chocking only the presence of vehicles at prespecified points(small area), whereas spatial information can be obtained by monitoring large area of traffic scene. To obtain spatial information by image processing, we need to track vehicles in the whole area of traffic scene. Image detector system based on global tracking consists of video input, vehicle detection, vehicle tracking, and traffic information measurement. For video input, conventional approaches used auto iris which is very poor in adaptation for sudden brightness change. Conventional methods for background generation do not yield good results in intersections with heave traffic and most of the early studies measure only point information. In this paper, we propose user-controlled iris method to remedy the deficiency of auto iris and design flame difference-based background generation method which performs far better in complicated intersections. We also propose measurement method for spatial traffic information such as interval volume/lime/velocity, queue length, and turning/forward traffic flow. We obtain measurement accuracy of 95%∼100% when applying above mentioned new methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37 no.7C
• /
• pp.554-564
• /
• 2012

The vehicle recognition consists of two steps; the vehicle region detection step and the vehicle identification step based on the feature extracted from the detected region. Features using linear transformations have the effect of dimension reduction as well as represent statistical characteristics, and show the robustness in translation and rotation of objects. Among the linear transformations, the NMF(Non-negative Matrix Factorization) is one of part-based representation. Therefore, we can extract NMF features with sparsity and improve the vehicle recognition rate by the representation of local features of a car as a basis vector. In this paper, we propose a feature extraction using NMF suitable for the vehicle recognition, and verify the recognition rate with it. Also, we compared the vehicle recognition rate for the occluded area using the SNMF(sparse NMF) which has basis vectors with constraint and LVQ2 neural network. We showed that the feature through the proposed NMF is robust in the urban scene where occlusions are frequently occur.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2018.10a
• /
• pp.497-498
• /
• 2018

In the fire scene, it is difficult to see 1m ahead because of power failure, smoke and toxic gas, even with thermal imaging camera and Xenon searchlight. Analysis of the smoke particles in the fire scene shows that even if the smoke is $5{\mu}m$ or less in wavelength, it is difficult to obtain a front view when using a conventional thermal imaging camera if the visual distance exceeds 1 meter. In the case of black smoke with a particle wavelength of $5{\mu}m$ or more, a space permeation sensor technology using various sensors other than a single sensor is required because chemical materials, gas, and water molecules are mixed. Firefighters need a smoke detection technology for smoke detection and spatial information visualization for forward safety view.In this paper, we design the interface of the information processing system with 32bit CPU core and peripheral circuit. We also implemented and simulated the interface with Lidar sensor. Through this, we provide interface that can implement information processing system of human enhancement fire helmet in the future.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.10C
• /
• pp.1414-1424
• /
• 2004

In this paper, an algorithm using wavelet transform for detecting a cut that is a radical scene transition point, and fade and dissolve that are gradual scene transition points is proposed. The conventional methods using wavelet transform for this purpose is using features in both spatial and frequency domain. But in the proposed algorithm, the color space of an input image is converted to YUV and then luminance component Y is transformed in frequency domain using 2-level lifting. Then, the histogram of only low frequency subband that may contain some spatial domain features is compared with the previous one. Edges obtained from other higher bands can be divided into global, semi-global and local regions and the histogram of each edge region is compared. The experimental results show the performance improvement of about 17% in recall and 18% in precision and also show a good performance in fade and dissolve detection.

• Journal of the Korea Society of Computer and Information
• /
• v.10 no.6 s.38
• /
• pp.65-74
• /
• 2005

This Paper Propose a robust scene change detection technique that use the weighted chi-square test and the automated threshold-decision algorithm. The weighted chi-test can subdivide the difference values of individual color channels by calculating the color intensities according to mSC standard, and it can detect the scene change by joining the weighted color intensities to the predefined chi-test which emphasize the comparative color difference values. The automated decision algorithm uses the difference values of frame-to-frame that was obtained by the weighted chi-test. In the first step, The average of total difference value and standard deviation value is calculated and then, subtract the mean value from the each difference values. In the next step, the same process is performed on the remained difference value. The propose method is tested on various sources and in the experimental results, it is shown that the Proposed method is efficiently estimates the thresholds and reliably detects scene changes.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.6
• /
• pp.1283-1288
• /
• 2011

In this paper we collect data concerning sleep environments in a bedroom and analyze the relationship between the collected condition data and sleep. In addition, this paper detects scene changes from the subjects in a sleeping state and presents the physical conditions, reactions during sleep, and physical sensations and stimuli. To detect scene changes in image sequences, we used color histogram for the difference between the preceding frame and the current frame. In addition, to extract the tossing and turning for different situations, the subjects were instructed to enter the level of fatigue, the level of drinking, and the level of stomach emptiness. For the sleep experiment system, we used the H-MOTE2420 Sensor composed of temperature, humidity, and light sensors. This paper is intended to provide the best sleep environment that enhances sleep quality, thus inducing people today to get regular and comfortable sleep.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.5
• /
• pp.35-44
• /
• 2004

With increase in digitalmedia data, various video indexing and video sequence matching algorithms have been proposed to efficiently manage and utilize digital media. In this paper, we propose a robust video indexing algorithm to detect scene changes for video sequences with abrupt luminance variations and an efficient video sequence matching algorithm for video sequence query. To improve the accuracy and to reduce the computational complexity for video indexing with abrupt luminance variations, the proposed algorithm utilizes edge features as well as color features, which yields a remarkably better performance than conventional algorithms. In the proposed algorithm first we extract the candidate shot boundaries using color histograms and then determine using edge matching and luminance compensation if they are shot boundaries or luminance changes. If the scene contains trivial brighness variations, the edge matching and luminance compensation are performed only for shot boundaries. In experimental results, the proposed method gives remarkably a high performance and efficiency than the conventional methods with the similar computational complexity.

• Journal of Broadcast Engineering
• /
• v.18 no.2
• /
• pp.185-195
• /
• 2013

In this paper, we propose a video summarization technique on a smartphone, based on spatiotemporal saliency. The proposed technique detects scene changes by computing the difference of the color histogram, which is robust to camera and object motion. Then the similarity between adjacent frames, face region, and frame saliency are computed to analyze the spatiotemporal saliency in a video clip. Over-segmented hierarchical tree is created using scene changes and is updated iteratively using mergence and maintenance energies computed during the analysis procedure. In the updated hierarchical tree, segmented frames are extracted by applying a greedy algorithm on the node with high saliency when it satisfies the reduction ratio and the minimum interval requested by the user. Experimental result shows that the proposed method summaries a 2 minute-length video in about 10 seconds on a commercial smartphone. The summarization quality is superior to the commercial video editing software, Muvee.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.5
• /
• pp.275-280
• /
• 2013

In this paper, a new stereo object extraction algorithm using a block-based MSE (mean square error) algorithm and the configuration parameters of a stereo camera is proposed. That is, by applying the SSD algorithm between the initial reference image and the next stereo input image, location coordinates of a target object in the right and left images are acquired and then with these values, the pan/tilt system is controlled. And using the moving angle of this pan/tilt system and the configulation parameters of the stereo camera system, the mask window size of a target object is adaptively determined. The newly segmented target image is used as a reference image in the next stage and it is automatically updated in the course of target tracking basing on the same procedure. Meanwhile, a target object is under tracking through continuously controlling the convergence and FOV by using the sequentiall extracted location coordinates of a moving target.