• Journal of Korea Multimedia Society
• /
• v.17 no.4
• /
• pp.420-432
• /
• 2014

In detection of moving objects from video sequences, an essential process for intelligent visual surveillance, the cast shadows accompanying moving objects are different from background so that they may be easily extracted as foreground object blobs, which causes errors in localization, segmentation, tracking and classification of objects. Most of the previous research results about moving cast shadow detection and removal usually utilize color information about objects and scenes. In this paper, we proposes a novel cast shadow removal method of moving objects in gray level video data for visual surveillance application. The proposed method utilizes observations about edge patterns in the shadow region in the current frame and the corresponding region in the background scene, and applies Laplacian edge detector to the blob regions in the current frame and the corresponding regions in the background scene. Then, the product of the outcomes of application determines moving object blob pixels from the blob pixels in the foreground mask. The minimal rectangle regions containing all blob pixles classified as moving object pixels are extracted. The proposed method is simple but turns out practically very effective for Adative Gaussian Mixture Model-based object detection of intelligent visual surveillance applications, which is verified through experiments.

• Journal of Communications and Networks
• /
• v.15 no.3
• /
• pp.283-291
• /
• 2013

In this paper, we propose new multiuser detectors (MUDs) based on compressed sensing approaches for the large-scale multiple antenna systems equipped with dozens of low-power antennas. We consider the scenarios where the number of receiver antennas is smaller than the total number of users, but the number of active users is relatively small. This prior information motivates sparsity-embracing MUDs such as sparsity-embracing linear/nonlinear MUDs where the detection of active users and their symbol detection are employed. In addition, sparsity-embracing MUDs with maximum a posteriori probability criterion (MAP-MUDs) are presented. They jointly detect active users and their symbols by exploiting the probability of user activity, and it can be solved efficiently by introducing convex relaxing senses. Furthermore, it is shown that sparsity-embracing MUDs exploiting common users' activity across multiple symbols, i.e., frame-by-frame, can be considered to improve performance. Also, in multiple multiple-input and multiple-output networks with aggressive frequency reuse, we propose the interference cancellation strategy for the proposed sparsity-embracing MUDs. That first cancels out the interference induced by adjacent networks and then recovers the desired users' information by exploiting the low user activity. In simulation studies for binary phase shift keying modulation, numerical evidences establish the effectiveness of our proposed MUDs exploiting low user activity, as compared with the conventional MUD.

• International Journal of Advanced Culture Technology
• /
• v.9 no.4
• /
• pp.260-267
• /
• 2021

In this paper, we intend to use a new mixed model of YoloV5 and DeepSort. For fire detection, we want to increase the accuracy by automatically extracting the characteristics of the flame in the image from the training data and using it. In addition, the high false alarm rate, which is a problem of fire detection, is to be solved by using this new mixed model. To confirm the results of this paper, we tested indoors and outdoors, respectively. Looking at the indoor test results, the accuracy of YoloV5 was 75% at 253Frame and 77% at 527Frame, and the YoloV5+DeepSort model showed the same accuracy at 75% at 253 frames and 77% at 527 frames. However, it was confirmed that the smoke and fire detection errors that appeared in YoloV5 disappeared. In addition, as a result of outdoor testing, the YoloV5 model had an accuracy of 75% in detecting fire, but an error in detecting a human face as smoke appeared. However, as a result of applying the YoloV5+DeepSort model, it appeared the same as YoloV5 with an accuracy of 75%, but it was confirmed that the false positive phenomenon disappeared.

• Annual Conference of KIPS
• /
• 2020.11a
• /
• pp.967-970
• /
• 2020

In a previous work we have done, we presented a monitoring system to automatically detect some dogs' behaviors from videos. However, the input video data used by that system was pre-trimmed to ensure it contained a dog only. In a real-life situation, the monitoring system would continuously receive video data, including frames that are empty and ones that contain people. In this paper, we propose a YOLOv4-based system for automatic object detection and trimming of dog videos. Sequences of frames trimmed from the video data received from the camera are analyzed to detect dogs and people frame by frame using a YOLOv4 model, and then records of the occurrences of dogs and people are generated. The records of each sequence are then analyzed through a rule-based decision tree to classify the sequence, forward it if it contains a dog only or ignore it otherwise. The results of the experiments on long untrimmed videos show that our proposed method manages an excellent detection performance reaching 0.97 in average of precision, recall and f-1 score at a detection rate of approximately 30 fps, guaranteeing with that real-time processing.

• Journal of Korea Multimedia Society
• /
• v.16 no.4
• /
• pp.430-443
• /
• 2013

The recognition of current position and road surrounding of car is very important to driver for safe driving. This paper presents the recognition technique of the road traveling environment using position information hiding and viewpoint transform that monitors the information of road furniture and signature and notifies them to driver. The proposed scheme generates the road images into which the position information are hided, from car camera and GPS module and provides the road information to driver through the viewpoint transformation and the road signature detection. The driving tests with camera and GPS module verified that the position information hiding takes about 66.5ms per frame, the detection rate of road signature is about 95.83%, and the road signature detection takes about 227.45ms per frame. Therefore, we know that the proposed scheme can recognize the road traveling environment on the road video with 15 frame rate.

• Journal of IKEEE
• /
• v.26 no.2
• /
• pp.265-270
• /
• 2022

This paper proposes the detection algorithm for switch open fault of asymmetric 6-phase PMSM based on stationary reference frame dq-axis currents. In this paper, target motor has an asymmetric structure in which two upper three windings have an electrical phase difference of 30° and a neutral point is separated. As a result, dual 3-phase PWM inverters and the detection techniques due to open failures of switch are definitely required. In this paper, the dual dq-axis current control method is used for driving the asymmetric 6-phase PMSM and the open fault switch should be detected by using variable all pass filter and low pass filter in order to detect the current amplitude. The effectiveness and usefulness of the proposed method is verified by several experiments.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.12
• /
• pp.141-149
• /
• 2013

This paper presents a high-speed intrusion detection process for multi-core video surveillance systems. The high-speed intrusion detection was designed to a parallel process. Based on the analysis of the conventional process, a parallel intrusion detection process was proposed so as to be accelerated by utilizing multiple processing cores in contemporary computing systems. The proposed process performs the intrusion detection in a per-frame parallel manner, considering the data dependency between frames. The proposed process was validated by implementing a multi-threaded intrusion detection program. For the system having eight processing cores, the detection speed of the proposed program is higher than that of the conventional one by up to 353.76% in terms of the frame rate.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.2
• /
• pp.365-380
• /
• 2022

The target detection algorithm based on supervised learning is the current mainstream algorithm for target detection. A high-quality dataset is the prerequisite for the target detection algorithm to obtain good detection performance. The larger the number and quality of the dataset, the stronger the generalization ability of the model, that is, the dataset determines the upper limit of the model learning. The convolutional neural network optimizes the network parameters in a strong supervision method. The error is calculated by comparing the predicted frame with the manually labeled real frame, and then the error is passed into the network for continuous optimization. Strongly supervised learning mainly relies on a large number of images as models for continuous learning, so the number and quality of images directly affect the results of learning. This paper proposes a dataset STAR-24K (meaning a dataset for Space TArget Recognition with more than 24,000 images) for detecting common targets in space. Since there is currently no publicly available dataset for space target detection, we extracted some pictures from a series of channels such as pictures and videos released by the official websites of NASA (National Aeronautics and Space Administration) and ESA (The European Space Agency) and expanded them to 24,451 pictures. We evaluate popular object detection algorithms to build a benchmark. Our STAR-24K dataset is publicly available at https://github.com/Zzz-zcy/STAR-24K.

• KIISE Transactions on Computing Practices
• /
• v.22 no.3
• /
• pp.133-138
• /
• 2016

Motion detection algorithms based on difference image are classified into background subtraction and previous frame subtraction. 1) Background subtraction is a convenient and effective method for detecting foreground objects in a stationary background. However in real world scenarios, especially outdoors, this restriction, (i.e., stationary background) often turns out to be impractical since the background may not be stable. 2) Previous frame subtraction is a simple technique for detecting motion in an image. The difference between two frames depends upon the amount of motion that occurs from one frame to the next. Both these straightforward methods fail when the object moves very "slightly and slowly". In order to efficiently deal with the problem, in this paper we present an algorithm for motion detection that incorporates "reflected light area" and "difference image". This reflected light area is generated during the frame production process. It processes multiplex difference image and AND-arithmetic of bitwise. This process incorporates the accuracy of background subtraction and environmental adaptability of previous frame subtraction and reduces noise generation. Also, the performance of the proposed method is demonstrated by the performance assessment of each method using Gait database sample of CASIA.

• The KIPS Transactions:PartB
• /
• v.10B no.6
• /
• pp.705-710
• /
• 2003

Recent years, as a rapid development of multimedia technology, video database system to retrieve video data efficiently seems to core technology in the oriented society. This thesis describes an efficient automatic frame detection and location method for content based retrieval of video. Frame extraction part is consist of incoming / outgoing car frame extraction and car number frame extraction stage. We gain star/end time of car video also car number frames. Frames are selected at fixed time interval from video and key frames are selected by color scale histogram and edge operation method. Car frame recognized can be searched by content based retrieval method.