• Journal of the Semiconductor & Display Technology
• /
• v.23 no.3
• /
• pp.135-139
• /
• 2024

Public datasets, which are freely available and often labeled, play a crucial role in training object detection models in computer vision. While public datasets are effective for developing general object detection models, they may not be ideal for specialized tasks. For specific object detection needs, it is more beneficial to create and use a dataset tailored to the target object. This paper proposes a method for extracting a target-specific dataset from public datasets to develop object detection models with superior performance for the target object. This approach not only improves detection accuracy, but also reduces training data requirements and complexity. We evaluate the performance of the proposed method using the latest object detection model YOLOv10.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.6C
• /
• pp.770-776
• /
• 2004

The goal of perceptual audio coder is to reduce redundancy and irrelevancy of audio signal based on the concept of masking. Several studies on masking effect reveal that the masking threshold varies as a function of the noise-like or tone-like nature of audio signals. Therefore, tonality of audio signal influences significantly the quality and efficiency of perceptual audio coder In this paper, we propose a new effective algorithm for tonality measure using spectrum energy. Since the proposed algorithm consists of a few transcendental functions and simple operations, it has lower complexity than MPEG psychoacoustic model-II. The proposed algorithm was tested with some audio signals, and DSP implementation showed that the proposed algorithm could be implemented with 3 MIPS. These results illustrate the efficiency of proposed algorithm in both performance and complexity.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.9 no.5
• /
• pp.205-211
• /
• 2009

The Weightless Neural Network(WNN) has the advantage of the processing speed, less computability than weighted neural network which readjusts the weight. Especially, The behavior information such as sequential gesture has many serial correlation. So, It is required the high computability and processing time to recognize. To solve these problem, Many algorithms used that added preprocessing and hardware interface device to reduce the computability and speed. In this paper, we proposed the Ram based Sequential Cumulative Neural Network(SCNN) model which is sign language recognition system without preprocessing and hardware interface. We experimented with using compound words in continuous korean sign language which was input binary image with edge detection from camera. The recognition system of sign language without preprocessing got 93% recognition rate.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.42 no.1
• /
• pp.33-42
• /
• 2005

Identifying color codes needs precise color information of their constituents, and is far from trivial because colors usually suffer severe distortions throughout the entire procedures from printing to acquiring image data. To accomplish accurate identification of colors, we need a reliable segmentation method to separate different color regions from each other, which would enable us to process the whole pixels in the region of a color statistically, instead of a subset of pixels in the region. Color image segmentation can be accomplished by performing edge detection on component image(s). In this paper, we separately detected edges on component images from RGB, HSI, and YIQ color models, and performed mathematical analyses and experiments to find out a pair of component images that provided the best edge image when combined. The best result was obtained by combining Y- and R-component edge images.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.13 no.1
• /
• pp.121-127
• /
• 2013

A vision based hand-gesture interface method for substituting a pointing device is proposed in this paper, which is used the ratio and size variation of Gesture Region. Proposed method uses the skin hue&saturation of the hand region from the HSI color model to extract the hand region effectively. This method can remove the non-hand region, and reduces the noise effect by the light source. Also, as the computation quantity is reduced by detecting not the static hand-shape recognition, but the ratio and size variation of hand-moving from the clipped hand region in real time, more response speed is guaranteed. In order to evaluate the performance of the our proposed method, after applying to the computerized self visual acuity testing system as a pointing device. As a result, the proposed method showed the average 86% gesture recognition ratio and 87% coordinate moving recognition ratio.

• Journal of the Korea Computer Industry Society
• /
• v.6 no.2
• /
• pp.357-364
• /
• 2005

The prefabricated type used generally in Korea to join cable runs on new installations or to repair broken Cable runs on existing installations, because installation is very simple and save time. This type is a permanent, shielded and submersible cable joint for direct burial or vault application. It confirms to the requirements of IEEE std. 404-1993 by factory testing, but many problems of insulated cable systems are caused by internal defects of the joint part which have to be mounted ensile. Faults arise from impurities or voids. A suitable solution for a monitoring of cable joints during the after-laying test and service is partial discharge detection. Specimen obtained 1mm thickness from the insulation of real power cable and cable joint. <중략>The partial discharges are measured to determine their time dependence for 60 minutes and the influence of applied electrical stress under 30kV. $\Phi-q-n$ properties were measured using detection impedance, high pass filter and computerized data acquisition system. Statistic Value like maximum charge, repetition rate, average charge, etc. are calculated. It is possible to quantitative analysis of $\Phi-q-n$ properties from this statistic value and pattern analysis.

• Journal of the Korea Computer Industry Society
• /
• v.6 no.2
• /
• pp.279-286
• /
• 2005

The prefabricated type used generally in Korea to join cable runs on new installations or to repair broken Cable runs on existing installations, because installation is very simple and save time. This type is a permanent, shielded and submersible cable joint for direct burial or vault application. It confirms to the requirements of IEEE std. 404-1993 by factory testing, but many problems of insulated cable systems are caused by internal defects of the joint part which have to be mounted ensile. Faults arise from impurities or voids. A suitable solution for a monitoring of cable joints during the after-laying test and service is partial discharge detection. <중략> $\Phi-q-n$ properties were measured using detection impedance, high pass filter and computerized data acquisition system. Statistic Value like maximum charge, repetition rate, average charge, etc. are calculated. It is possible to quantitative analysis of $\Phi-q-n$ properties from this statistic value and pattern analysis.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.12 no.1
• /
• pp.151-158
• /
• 1992

Topographic mapping from economic SPOT stereo imagery than aerial photographs has become possible. Many of authors have studied the possibility of base map revisions and the accuracy of results. They have concluded that the SPOT image is suitable for 1/50,000 to 1/100,000 topographic map. For topographic map, orthophoto and DTM generation from SPOT imagery, accurate exterior orientation parameters are needed. But since the geometric characteristic of SPOT image is dynamic linear array imagery, the conventional bundle adjustment for photogrammetry can not be directly applied. Reliability is the ability to detect gross error, which is called the internal reliability, and the effect of non-detectable gross error on the results of exterior orientation, which is called the external reliability. This paper shows how the reliability of SPOT imagery depends on the different coordinate systems, presentations of coordinate for flight direction, orders of exterior orientation parameters and distribution of control points, and thus analyses the theoretical reliability of the exterior orientation, which can provide a basis for the planning of SPOT projects.

• Progress in Medical Physics
• /
• v.15 no.2
• /
• pp.112-119
• /
• 2004

We tried to build human models based on medical images of live Korean, instead of using standard data of human body structures. Characteristics of mechanical structures of human bodies were obtained from medical images such as CT and MR images. For each constitutional part of mechanical structures CT images were analyzed in terms of gray levels and MR images were analyzed in terms of pulse sequence. Characteristic features of various mechanical structures were extracted from the analyses. Based on the characteristics of each structuring element we peformed image segmentation on CT and MR images. We delineated bones, muscles, ligaments and tendons from CT and MR images using image segmentation or manual drawing. For the image segmentation we compared the edge detection method, region growing method and intensity threshold method and applied an optimal compound of these methods for the best segmentation results. Segmented mechanical structures of the head/neck part were three dimensionally reconstructed.

• Journal of Digital Convergence
• /
• v.12 no.6
• /
• pp.231-236
• /
• 2014

In this paper, we propose a processing speed improvement by adding a parallel processing based on device(graphic card) into a road region extraction by host(PC) based serial processing. The OpenCV CUDA supports the many functions of parallel processing method by interworking a conventional OpenCV with CUDA. Also, when interworking the OpenCV and CUDA, OpenCV functions completed a configuration are optimized the User's device(Graphic Card) specifications. Thus, OpenCV CUDA usage provides an algorithm verification and easiness of simulation result deduction. The proposed method is verified that the proposed method has a about 3.09 times faster processing speed than a conventional method by using OpenCV CUDA and graphic card of NVIDIA GeForce GTX 560 Ti model through experimentation.