• Journal of the Society of Naval Architects of Korea
• /
• v.56 no.6
• /
• pp.473-479
• /
• 2019

Welding is one of the most popular joining methods and most welding quality estimation methods are executed using joined material. This paper propose welding quality estimation methods using dynamic current, voltage and resistance which are obtained during welding in real time. There are many kinds of welding method. Among them, we focused on the projection welding and gathered dynamic characteristics from two different types of projection welding. For image learning, graphs are drawn using obtained current, voltage and resistance, and the graphs are converted to images. The images are labeled with two sub-categories - normal and defect. For deep learning of images obtained from welding, Convolutional Neural Network (CNN) is applied, and Tensorflow was used as a framework for deep learning. With two resistance welding test datasets, we conclude that the Convolutional Neural Network helps in predicting the welding quality.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.65 no.8
• /
• pp.1407-1417
• /
• 2016

In this paper, we propose a multi-level content transmission mechanism for intelligent quality of service (QoS) in social networking services (SNSs). Because existing SNSs and related work send image content to a client with a single fixed mechanism, they cannot consistently support content accessibility according to different conditions of QoS factors such as network congestion and throughput. In the proposed image transmission mechanism, our communication middleware (CM) provides an SNS developer with three transmission modes so that an SNS server or client can dynamically change the quality of images if required. In each transmission mode, an SNS server can send images to a requesting client with original high quality, thumbnail quality, or send only text information. With varying qualities of downloaded images, an SNS developed on top of CM can provide users with consistent QoS for access to SNS content.

• Journal of the Institute of Convergence Signal Processing
• /
• v.11 no.4
• /
• pp.278-288
• /
• 2010

In this paper, we carried out quality assessment and analysis of stereoscopic 3D Television pictures according to ITU-R contribution and recommendation by rating scales using DSCQS (Double-Stimuli Continuous Quality Scale) method. The evaluation results show that overall quality and sharpness of stereoscopic pictures revealed almost no difference compared to cases of mono pictures as to natural outdoor scenes, graphic images, and indoor scenes (about 3.0 above ~ 4.0), but depth perception and sensation of reality of stereoscopic images exhibit better quality performance over mono images as indicated more than 4.0 out of 5.0 grade. Evaluation results should be considered as human factors such as disparity when shooting and/or editing 3DTV.

• Korean Journal of Human Ecology
• /
• v.8 no.1
• /
• pp.5-18
• /
• 2005

The purposes of this study were to identify brand image and advertising image perception maps of domestic make-up products. A self-administered questionnaire was used for data collection. KYST, CORAN and SPSS PC(Ver. 12.0) were used for data analysis. The results were as follows: 1. Brand images of Etude, Isa Knox, and Laneige were perceived as unique, stimulative, high quality, elegant, modern, and sophisticated. Brand image of Cathycat was perceived highly in high quality, elegant, modern, and sophisticated, but low in unique and stimulative. Brand image of Lac Vert was perceived high in unique and stimulative, but low in high quality, elegant, modern, and sophisticated. Brand images of Hercyna and Vov were the lowest. 2. Advertisement images of Etude was perceived as modern, sophisticated, familiar, and unique, but Lac Vert was perceived adversely, Advertising images of Laneigne and Isa Knox were high in modern, sophisticated and familiar, but low in uniqueness. And advertising images of Hercyna, Cathycat, and Vov were perceived as modern, sophisticated, and familiar.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.25 no.6
• /
• pp.663-671
• /
• 1988

In this paper, in order to reduce the blocking effect of block coded images, we propose the method considering the lowpass and bandpass components of Granrath's human visual model. This method consists of two-stage enhancement procedure. The first step is lowpass filtering which smooths out the blocking effect, and the second step is a high frequency enhancement procedure to increase the contrast decreased by the lowpass filtering in the first step. In the first step, the one-dimensional Gaussian filter which aligthns parallel to the edge direction is considered to preserve the edge in the block and the two-dimensional Gaussian filter is used to smooth out the blocking effect near the block boundaries. In the second step, the lowpass and bandpass components of the Granrath's model are considered to increase contrast in a restored image. The performance comparison of the proposed method and the existing mehtods is made by a computer simulation with several block coded images. We can see that the enhancement in the subjective quality of images of the proposed method is more significant than the enhancement in the subjective quality of images of the proposed method is more significant than the existing methods, though the proposed method does not show better performance on the PSNR gain, the poor measure of picture quality for block coded images.

• Journal of Korean Academy of Oral and Maxillofacial Radiology
• /
• v.23 no.1
• /
• pp.125-136
• /
• 1993

The geometrically standardized intraoral radiographs using 5 occlusal registration materials were taken serially from immediate, 1 day, 2, 4, 8, 12, and 16 weeks after making the bite blocks. And the resultant images were digitally subtracted using the immediately taken film as reference images. The qualities of those subtracted images were evaluated to check the degree of reproducibility of each impression material. The results were as follows: 1. The standard deviations of the grey scales of the overall subtracted images were 4.9 for Exal1ex, 7.2 for Pattern resin, 9.0 for Tooth Shade Acrylic, 12.2 for XCP only, 14.8 for Impregum. the lesser the standard deviation, the better the quality of the subtracted images. 2. The standard deviation of the grey scales of the overall subtracted images were grossly related to those of the localized horizontal line of interest. 3. Exaflex which showed the best subtracted image quality had 15 cases of straight, 14 cases of wave, 1 case of canyon shape. Impregum which showed the worst subtracted image quality had 4 cases of straight, 8 cases of wave, 18 cases of canyon shape respectively.

• Smart Media Journal
• /
• v.6 no.3
• /
• pp.57-63
• /
• 2017

The most important, essential prerequisite for the improvement of vehicle license plate recognition is the acquisition of high-quality vehicle images. Because typical images acquired from roads are affected by different environmental factors including the time of day, sunlight, and the weather, the brightness and the shape of the license plates in the images are inconsistent. To this end, many image corrections are performed, resulting in slower recognition and lower recognition rate. Therefore, in this study, we used the images acquired from roads to test the proposed method for fast capturing of vivid, high-quality vehicle images by measuring the brightness around license plates during real-time image capturing to control in real time the factors, such as shutter speed, brightness, and gain of the camera, that affect the brightness and the quality of the images.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.18 no.4
• /
• pp.1-11
• /
• 2022

Recently, research on license plate recognition, which is a core technology of an intelligent transportation system(ITS), is being actively conducted. In this paper, we propose a method to extract numbers and characters from low-quality license plate images by applying the YOLOv4 algorithm. YOLOv4 is a one-stage object detection method using convolution neural network including BACKBONE, NECK, and HEAD parts. It is a method of detecting objects in real time rather than the previous two-stage object detection method such as the faster R-CNN. In this paper, we studied a method to directly extract number and character regions from low-quality license plate images without additional edge detection and image segmentation processes. In order to evaluate the performance of the proposed method we experimented with 500 license plate images. In this experiment, 350 images were used for training and the remaining 150 images were used for the testing process. Computer simulations show that the mean average precision of detecting number and character regions on vehicle license plates was about 93.8%.

• International Journal of Computer Science & Network Security
• /
• v.21 no.9
• /
• pp.275-280
• /
• 2021

With the development of communication networks, the processes of exchanging and transmitting information rapidly developed. As millions of images are sent via social media every day, also wireless sensor networks are now used in all applications to capture images such as those used in traffic lights, roads and malls. Therefore, there is a need to reduce the size of these images while maintaining an acceptable degree of quality. In this paper, we use Python software to apply K-mean Clustering algorithm to compress RGB images. The PSNR, MSE, and SSIM are utilized to measure the image quality after image compression. The results of compression reduced the image size to nearly half the size of the original images using k = 64. In the SSIM measure, the higher the K, the greater the similarity between the two images which is a good indicator to a significant reduction in image size. Our proposed compression technique powered by the K-Mean clustering algorithm is useful for compressing images and reducing the size of images.

• Journal of information and communication convergence engineering
• /
• v.19 no.3
• /
• pp.180-187
• /
• 2021

In this paper, we present a computational volumetric reconstruction method for three-dimensional (3D) photon counting imaging with enhanced visual quality when low-resolution elemental images are used under photon-starved conditions. In conventional photon counting imaging with low-resolution elemental images, it may be difficult to estimate the 3D scene correctly because of a lack of scene information. In addition, the reconstructed 3D images may be blurred because volumetric computational reconstruction has an averaging effect. In contrast, with our method, the pixels of the elemental image rearrangement technique and a Bayesian approach are used as the reconstruction and estimation methods, respectively. Therefore, our method can enhance the visual quality and estimation accuracy of the reconstructed 3D images because it does not have an averaging effect and uses prior information about the 3D scene. To validate our technique, we performed optical experiments and demonstrated the reconstruction results.