• Journal of IKEEE
• /
• v.23 no.3
• /
• pp.777-784
• /
• 2019

A pre-fire awareness and automatic notification system are required because it is possible to minimize the damage if the fire situation is precisely detected after a fire occurs in a place where people are unusual or in a mountainous area. In this study, we developed a RaspberryPi-based fire recognition system using Faster-recurrent convolutional neural network (F-RCNN) and single shot multibox detector (SSD) and demonstrated a fire alarm system that works with power line communication. Image recognition was performed with a pie camera of RaspberryPi, and the detected fire image was transmitted to a monitoring PC through an inductive power line communication network. The frame rate per second (fps) for each learning model was 0.05 fps for Faster-RCNN and 1.4 fps for SSD. SSD was 28 times faster than F-RCNN.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.47 no.6
• /
• pp.97-106
• /
• 2010

In this paper, lane and curvature detection algorithm based on the curve template matching method is proposed. To eliminate the perspective effect of the original image, the input image is transformed to a top view image. From this top view image, its edge image is created. To increase the accuracy of detection, a novel edge detection method, which shows a strength in lane detection, is proposed. In the first step, straight lanes are detected from the edge image, and then the Curve Template Matching(CTM) method is applied to detect the curved lanes and to find their curvatures. Since the proposed CTM method uses only the simple equations, such as line and circle equations, to detect the curved lane, the algorithm is simple. Moreover, we used the detected lane information in the previous frames to detect the current frame's lanes, the detection results become more reliable. The proposed algorithm has been tested in various road conditions (highway, urban street, night time highway, etc.). Experimental results show that the proposed algorithm can process about 70 frames per second with the successful lane detection rate over 95% and curvature detection rate about 90%.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.1
• /
• pp.15-31
• /
• 2013

This paper proposes an intermediate image generation method for the viewer's view point by tracking the viewer's face, which is converted to a digital hologram. Its purpose is to increase the viewing angle of a digital hologram, which is gathering higher and higher interest these days. The method assumes that the image information for the leftmost and the rightmost view points within the viewing angle to be controlled are given. It uses a stereo-matching method between the leftmost and the rightmost depth images to obtain the pseudo-disparity increment per depth value. With this increment, the positional informations from both the leftmost view point and the rightmost view point are generated, which are blended to get the information at the wanted intermediate viewpoint. The occurrable dis-occlusion region in this case is defined and a inpainting method is proposed. The results from implementing and experimenting this method showed that the average image qualities of the generated depth and RGB image were 33.83[dB] and 29.5[dB], respectively, and the average execution time was 250[ms] per frame. Also, we propose a prototype system to service digital hologram interactively to the viewer by using the proposed intermediate view generation method. It includes the operations of data acquisition for the leftmost and the rightmost viewpoints, camera calibration and image rectification, intermediate view image generation, computer-generated hologram (CGH) generation, and reconstruction of the hologram image. This system is implemented in the LabView(R) environments, in which CGH generation and hologram image reconstruction are implemented with GPGPUs, while others are implemented in software. The implemented system showed the execution speed to process about 5 frames per second.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.8
• /
• pp.1486-1491
• /
• 2017

In this paper, we propose a private multimedia caching scheme based on wireless local area network (WLAN) to improve the quality of service for high capacity and high quality multimedia streaming services which are recently increasing and to reduce the traffic load of core networks. The proposed caching scheme stores multimedia in the storage device mounted on WLAN APs and provides streaming services on its own without Internet connection in accordance with the request from clients. We have implemented a test network based on real commercial networks and measured the performance of the proposed caching scheme in terms of frames per second (FPS) and buffering time. According to the performance measurement results, the proposed caching scheme can reduce the average buffering time by 73.3% compared to the conventional streaming scheme. In addition, the proposed caching scheme can also improve the average FPS by 71.3% compared to the conventional streaming scheme.

• The KIPS Transactions:PartA
• /
• v.12A no.4 s.94
• /
• pp.327-332
• /
• 2005

In the field of computer graphics, we have several research results to display the ocean waves on the screen, while we still not have a complete solution yet. Though ocean waves are constructed from a variety of sources, the dominant one is the surface gravity wave, which is generated by the gravity and the wind. In this Paper, we Present a real-time surface gravity wave simulation method, derived from a precise ocean wave model in the oceanography. There are research results based on the Pierson-Moskowitz(PM) model[1], which assumes infinite depth of water and thus shows some mismatches in the case of shallow seas. In this paper, we started from the Texel, Marsen and Arsloe(TMA) model[2], which is a more precise wave model and thus can be used to display more realistic ocean waves. We derived its implementation model for the graphics applications and our prototype implementation shows about 30 frames per second on the Intel Pentium 4 1.6GHz-based personal computer. Our major contributions to the computer graphics area ill be (1) providing more user-controllable parameters to finally generate various wave shapes and (2) the improvement on the expression power of waves even in the shallow seas.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.6B
• /
• pp.930-939
• /
• 2010

Recently, many researches for making more comfortable input device based on gaze detection have been vigorously performed in human computer interaction. However, these previous researches are difficult to be used in IPTV environment because these methods need additional wearing devices or do not work at a distance. To overcome these problems, we propose a new way of controlling IPTV interface by using a detected face and eye positions in single static camera. And although face or eyes are not detected successfully by using Adaboost algorithm, we can control IPTV interface by using motion vectors calculated by pyramidal KLT (Kanade-Lucas-Tomasi) feature tracker. These are two novelties of our research compared to previous works. This research has following advantages. Different from previous research, the proposed method can be used at a distance about 2m. Since the proposed method does not require a user to wear additional equipments, there is no limitation of face movement and it has high convenience. Experimental results showed that the proposed method could be operated at real-time speed of 15 frames per second. Wd confirmed that the previous input device could be sufficiently replaced by the proposed method.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.44 no.6
• /
• pp.35-41
• /
• 2007

We proposed an eye gaze tracking system under natural head movements. The system consists of one CCD(charge-coupled device) camera and two front-surface mirrors. The mirrors rotate to follow head movements in order to keep the eye within the view of the camera. However, the mirror controller cannot guarantee the fast head movements, because the frame rate is generally 30Hz. To overcome this problem, we applied Kalman filter to estimate next eye position from the current eye image. In the results, our system allowed the subjects head to move 60cm horizontally and 40cm vertically, with the head movement speed about 55cm/sec and 45cm/sec, respectively. And spatial gate resolutions were about 4.5 degree and 5.0 degree, respectively, and the gaze estimation accuracy was 92% under natural head movements.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.3C
• /
• pp.296-306
• /
• 2003

In this paper, algorithm for non-iterative decoding method is proposed and fractal image decoder based on non-iterative fractal decoding algorithm used general purpose digital signal processors is designed and implemented. The algorithm is showed that the attractor image can be obtained analytically whe n the image is encoded using the fractal algorithm proposed by Monro and Dudbridge, in which the corresponding domain block for a range block is fifed. Using the analytical formulas, we can obtain the attractor image without iteration procedure. And we get general formulas of obtained analytical formulas. Computer simulation results for various test images show that we can increase the image decoding speed by more than five times when we use the analytical formulas compared to the previous iteration methods. The fractal image decoder contains two ADSP2181's and perform image decoding by three stage pipeline structure. The performance tests of the implemented decoder is elapsed 31.2ms/frame decoding speed for QCIF data when all the frames are decoded. The results enable us to process the real-time fractal decoding over 30 frames/sec.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.8C
• /
• pp.1113-1124
• /
• 2004

In this paper, we proposed a hardware(H/W) structure which can compress and recontruct the input image in real time operation and implemented it into a FPGA platform using VHDL(VHSIC Hardware Description Language). All the image processing element to process both compression and reconstruction in a FPGA were considered each of them was mapped into H/W with the efficient structure for FPGA. We used the DWT(discrete wavelet transform) which transforms the data from spatial domain to the frequency domain, because use considered the motion JPEG2000 as the application. The implemented H/W is separated to both the data path part and the control part. The data path part consisted of the image processing blocks and the data processing blocks. The image processing blocks consisted of the DWT Kernel fur the filtering by DWT, Quantizer/Huffman Encoder, Inverse Adder/Buffer for adding the low frequency coefficient to the high frequency one in the inverse DWT operation, and Huffman Decoder. Also there existed the interface blocks for communicating with the external application environments and the timing blocks for buffering between the internal blocks The global operations of the designed H/W are the image compression and the reconstruction, and it is operated by the unit of a field synchronized with the A/D converter. The implemented H/W used the 69%(16980) LAB(Logic Array Block) and 9%(28352) ESB(Embedded System Block) in the APEX20KC EP20K600CB652-7 FPGA chip of ALTERA, and stably operated in the 70MHz clock frequency. So we verified the real time operation of 60 fields/sec(30 frames/sec).

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.7A
• /
• pp.683-694
• /
• 2002

In this paper, a digital image compression codec using 2D DWT(Discrete Wavelet Transform) is designed using the FPGA technology for real time operation The implemented image compression codec using wavelet decomposition consists of a wavelet kernel part for wavelet filtering process, a quantizer/huffman coder for quantization and huffman encoding of wavelet coefficients, a memory controller for interface with external memories, a input interface to process image pixels from A/D converter, a output interface for reconstructing huffman codes, which has irregular bit size, into 32-bit data having regular size data, a memory-kernel buffer to arrage data for real time process, a PCI interface part, and some modules for setting timing between each modules. Since the memory mapping method which converts read process of column-direction into read process of the row-direction is used, the read process in the vertical-direction wavelet decomposition is very efficiently processed. Global operation of wavelet codec is synchronized with the field signal of A/D converter. The global hardware process pipeline operation as the unit of field and each field and each field operation is classified as decomposition levels of wavelet transform. The implemented hardware used FPGA hardware resource of 11119(45%) LAB and 28352(9％) ESB in FPGA device of APEX20KC EP20k600CB652-7 and mapped into one FPGA without additional external logic. Also it can process 33 frames(66 fields) per second, so real-time image compression is possible.