• IEIE Transactions on Smart Processing and Computing
• /
• v.1 no.3
• /
• pp.125-132
• /
• 2012

This paper proposes an automatic object segmentation and background composition method for video communication over consumer mobile phones. The object regions were extracted based on the motion and color variance of the first two frames. To combine the motion and variance information, the Euclidean distance between the motion boundary pixel and the neighboring color variance edge pixels was calculated, and the nearest edge pixel was labeled to the object boundary. The labeling results were refined using the morphology for a more accurate and natural-looking boundary. The grow-cut segmentation algorithm begins in the expanded label map, where the inner and outer boundary belongs to the foreground and background, respectively. The segmented object region and a new background image stored a priori in the mobile phone was then composed. In the background composition process, the background motion was measured using the optical-flow, and the final result was synthesized by accurately locating the object region according to the motion information. This study can be considered an extended, improved version of the existing background composition algorithm by considering motion information in a video. The proposed segmentation algorithm reduces the computational complexity significantly by choosing the minimum resolution at each segmentation step. The experimental results showed that the proposed algorithm can generate a fast, accurate and natural-looking background composition.

• Journal of Information Processing Systems
• /
• v.18 no.2
• /
• pp.197-207
• /
• 2022

The applications and user interfaces (UIs) of smart mobile devices are constantly diversifying. For example, deep learning can be an innovative solution to classify widgets in screen images for increasing convenience. To this end, the present research leverages captured images and the ReDraw dataset to write deep learning datasets for image classification purposes. First, as the validation for datasets using ResNet50 and EfficientNet, the experiments show that the dataset composed in this study is helpful for classification according to a widget's functionality. An implementation for widget detection and classification on RetinaNet and EfficientNet is then executed. Finally, the research suggests the Widg-C and Widg-D datasets-a deep learning dataset for identifying the widgets of smart devices-and implementing them for use with representative convolutional neural network models.

• Journal of the Institute of Convergence Signal Processing
• /
• v.8 no.3
• /
• pp.163-170
• /
• 2007

In this paper, we propose that the polyphase scaler whose performance to that of the bicubic method, has less complexity in hardware structure. In order to get the new information, proposed system is based on the group delay which is one of the digital filter characteristics. The performance of this system is superior to that of bicubic algorithm which is well known. Because the hardware structure is simpler than other image scalers, we can adopt this system for mobile devices easily. The previous polyphase filters make blurring noise which is generated by up-scaling. We replace polyphase filters by boost-up filter to get vivid image. The proposed scaler is verified by Xilinx Virtex2 FPGA and is used as digital Boom in mobile camera phone.

• IEIE Transactions on Smart Processing and Computing
• /
• v.5 no.6
• /
• pp.383-389
• /
• 2016

This paper presents a novel automatic white balance (AWB) algorithm for consumer imaging devices. While existing AWB methods require reference white patches to correct color, the proposed method performs the AWB function using only an input image in two steps: i) white point detection, and ii) color constancy gain computation. Based on the dark channel prior assumption, a white point or region can be accurately extracted, because the intensity of a sufficiently bright achromatic region is higher than that of other regions in all color channels. In order to finally correct the color, the proposed method computes color constancy gain values based on the Y component in the XYZ color space. Experimental results show that the proposed method gives better color-corrected images than recent existing methods. Moreover, the proposed method is suitable for real-time implementation, since it does not need a frame memory for iterative optimization. As a result, it can be applied to various consumer imaging devices, including mobile phone cameras, compact digital cameras, and computational cameras with coded color.

• Journal of the Institute of Convergence Signal Processing
• /
• v.16 no.1
• /
• pp.22-28
• /
• 2015

In this paper, a service system which shows linked video clips from the static images extracted from newspapers, magazines, photo albums and etc in an augmented reality. First, the system uses SURF algorithm to extract features from the original photos printed in the media and stores them with the linked video clips. Next, when a photo is taken by using a camera from mobile devices such as smart phones, the system extracts features in real time, search a linked video clip matching the original image, and shows it on the smart phone in an augmented reality. The proposed system is applied to Android smart phone devices and the test results verify that the proposed system operates not only on normal photos but also on partially damaged photos.

• Journal of the Optical Society of Korea
• /
• v.10 no.3
• /
• pp.124-129
• /
• 2006

We have developed a compact and cost-effective camera module on the basis of wafer-scale-replica processing. A multiple-layered structure of several aspheric lenses in a mobile-phone camera module is first assembled by bonding multiple glass-wafers on which 2-dimensional replica arrays of identical aspheric lenses are UV-embossed, followed by dicing the stacked wafers and packaging them with image sensor chips. This wafer-scale processing leads to at least 95% yield in mass-production, and potentially to a very slim phone with camera-module less than 2 mm in thickness. We have demonstrated a VGA camera module fabricated by the wafer-scale-replica processing with various UV-curable polymers having refractive indices between 1.4 and 1.6, and with three different glass-wafers of which both surfaces are embossed as aspheric lenses having $230{\mu}m$ sag-height and aspheric-coefficients of lens polynomials up to tenth-order. We have found that precise compensation in material shrinkage of the polymer materials is one of the most technical challenges, in orderto achieve a higher resolution in wafer-scaled lenses for mobile-phone camera modules.

• Journal of KIISE:Computing Practices and Letters
• /
• v.10 no.4
• /
• pp.328-336
• /
• 2004

Recently a mobile service is changing into a system environment that offers the customer various contents service. Representative example of service is a Calling Identity Del ivory Service(CID). Such service has the problem in the case which the receiver cannot remember the phone number of the sender; it has a difficult problem that cannot easily confirm whose the phone number it is. Therefore, it is desirable to design and implement visual services that can enhance the recognition of users. In this paper, we propose the architecture that is similar to a Calling Identity Delivery Service. We propose the architecture for communication service and system that is able to visually display the information of the sender using 2D image data in mobile environment. After that we set the image information to represent the user and this method is able to visually display the information of the sender by transmitting an image data through channels from switch station or base station using the server. When the receiver receives a phone call from the sender, this method provides an efficient service by transmitting visual data with bell sound. That is, the image information of sender is appeared on liquid crystal display(LCD) of the receiver at the same time. We investigate the concepts for processing real-time transmission of image data and describe an example of the implementation result that is based on system. This technology has a potential influence on the marketing and presents an efficiency of this method.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.24 no.5
• /
• pp.542-547
• /
• 2014

This paper describes the medical and IT convergence system using a smart phone and a heuristic method for the measurement of the dripping speed of the liquid in a drip chamber, which can estimate the remaining time using pattern recognition and difference image from video frame information based on Android technology. The video frames were first made using a smartphone camera and we calculated the difference image between the n image and the (n-1) image and then changed into binary images using the threshold value. At this point, it is very important to find an optimal threshold value using heuristic method to recognize the dripping of the liquids. In addition, the user can adjust the dripping speed according to the doctor's prescription, exactly like watching the progress bar of a mobile application. The experiment results show that our method using video processing technique accurately measures the dripping speed for a wide range of speeds that are sufficient for ordinary practice.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.11 no.3
• /
• pp.211-216
• /
• 2011

The detection of defective parts in a factory is usually performed by the human eye. Therefore, heavy manpower is in demand for minor enterprises. An image processing system is desired to solve this drawback. However, due to the variety of the products characteristics, an general algorithm is needed that can adapt to these characteristics. Therefore, in this paper, the key pad parts' characteristics which need to be dealt with are analyzed in order to embody the image processing algorithm that is suggested. The experimental results show the probability of detecting a defective part is 95% with a detection time of 0.203 seconds, on the average.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.169-171
• /
• 2021

Image processing through cameras such as self-driving, CCTV, mobile phone security, and parking facilities is being used to solve many real-life problems. Simple classification is solved through image processing, but it is difficult to find images or in-image features of complexly mixed objects. To solve this feature point, we utilize deep learning techniques in classification, detection, and segmentation of image data so that we can think and judge closely. Of course, the results are better than just image processing, but we confirm that the results judged by the method of image segmentation using deep learning have deviations from the real object. In this paper, we study how to perform accuracy improvement through simple image processing just before outputting the output of deep learning image segmentation to increase the precision of image segmentation.