• Current Optics and Photonics
• /
• v.2 no.2
• /
• pp.165-171
• /
• 2018

Fourier ptychographic microscopy (FPM) is a recently proposed computational imaging method that achieves both high resolution (HR) and wide field of view. In the FPM framework, a series of low-resolution (LR) images at different illumination angles is used for high-resolution image reconstruction. On the basis of previous research, image noise can significantly degrade the FPM reconstruction result. Since the captured LR images contain a lot of dark-field images with low signal-to-noise ratio, it is very important to apply a noise-reduction process to the FPM raw dataset. However, the thresholding method commonly used for the FPM data preprocessing cannot separate signals from background noise effectively. In this work, we propose an improved thresholding method that provides a reliable background-noise threshold for noise reduction. Experimental results show that the proposed method is more efficient and robust than the conventional thresholding method.

• Current Optics and Photonics
• /
• v.5 no.3
• /
• pp.298-305
• /
• 2021

Modern distance sensing methods employ various measurement principles, including triangulation, time-of-flight, confocal, interferometric and frequency comb. Among them, the triangulation method, with a laser light source and an image sensor, is widely used in low-cost applications. We developed an omnidirectional two-dimensional (2D) distance sensor based on the triangulation principle for indoor floor mapping applications. The sensor has a range of 150-1500 mm with a relative resolution better than 4% over the range and 1% at 1 meter distance. It rotationally scans a compact one-dimensional (1D) distance sensor, composed of a near infrared (NIR) laser diode, a folding mirror, an imaging lens, and an image detector. We designed the sensor layout and configuration to satisfy the required measurement range and resolution, selecting easily available components in a special effort to reduce cost. We built a prototype and tested it with seven representative indoor wall specimens (white wallpaper, gray wallpaper, black wallpaper, furniture wood, black leather, brown leather, and white plastic) in a typical indoor illuminated condition, 200 lux, on a floor under ceiling mounted fluorescent lamps. We confirmed the proposed sensor provided reliable distance reading of all the specimens over the required measurement range (150-1500 mm) with a measurement resolution of 4% overall and 1% at 1 meter, regardless of illumination conditions.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.6 no.4
• /
• pp.109-121
• /
• 2003

The point of this research is the use of the high resolution satellite image for local government's new address system, as well as spatially field investigation support and base map error finding. Most local governments use scale 1/1,000 and 1/5,000 digital map for base map and field investigation. But field investigator's knowledge insufficiency and the lack of base map's currency make things too difficult from the beginning of the project. As the way of solving this problem, this research offers the use of the high resolution satellite image in new address system with cadence data of digital base map. Until now satellite image is not suitable for our situation because it has low resolution. But this problem was solved for 1m space resolution satellite image and it is being applied wider and wider. Now vector data and Raster data are integrated for complimenting of each weak point. In this study the use of the high resolution satellite image in new address system is expected to improve the quality of the results and reduce the expenses. In addition the satellite image can use local government's fundamental data.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.7
• /
• pp.100-106
• /
• 2015

In this paper, we propose a fast algorithm for low-power image enhancement method for mobile LCD. In the proposed fast algorithm, the spatial resolution of the input image is significantly reduced, and the image characteristics are analyzed on the reduced resolution image to find a dimming rate adaptive to the image content, thereby saving power. The proposed fast adaptive dimming and image enhancement algorithm is implemented as an application that runs on an Android device. Image quality evaluation and running time analysis experiments on the device indicate that the proposed fast algorithm jointly minimizes the quality degradation and power consumption, reducing the required computation load by over 95%.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.48 no.2
• /
• pp.20-27
• /
• 2011

A CMOS Image Sensor(CIS) mounted on mobile appliances always needs a low power consumption because of the battery life cycle. In this paper, we propose novel power reduction techniques such as a data flip-flop circuit with leakage current elimination, a low power single slope A/D converter with a novel comparator, and etc. Based on 0.13um CMOS process, the chip satisfies QVGA resolution($320{\times}240$ pixels) whose pitch is 2.25um and whose structure is 4-Tr active pixel sensor. From the experimental results, the ADC in the middle of CIS has a 10-b resolution, the operating speed of CIS is 16 frame/s, and the power dissipation is 25mW at 3.3V(Analog)/1.8V(Digital) power supply. When we compare the proposed CIS with conventional ones, the power consumption is reduced approximately by 22% in sleep mode, 20% in operating mode.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.8
• /
• pp.2068-2082
• /
• 2023

With the rapid development of deep learning, Depth Map Super-Resolution (DMSR) method has achieved more advanced performances. However, when the upsampling rate is very large, it is difficult to capture the structural consistency between color features and depth features by these DMSR methods. Therefore, we propose a color-image guided DMSR method based on iterative depth feature enhancement. Considering the feature difference between high-quality color features and low-quality depth features, we propose to decompose the depth features into High-Frequency (HF) and Low-Frequency (LF) components. Due to structural homogeneity of depth HF components and HF color features, only HF color features are used to enhance the depth HF features without using the LF color features. Before the HF and LF depth feature decomposition, the LF component of the previous depth decomposition and the updated HF component are combined together. After decomposing and reorganizing recursively-updated features, we combine all the depth LF features with the final updated depth HF features to obtain the enhanced-depth features. Next, the enhanced-depth features are input into the multistage depth map fusion reconstruction block, in which the cross enhancement module is introduced into the reconstruction block to fully mine the spatial correlation of depth map by interleaving various features between different convolution groups. Experimental results can show that the two objective assessments of root mean square error and mean absolute deviation of the proposed method are superior to those of many latest DMSR methods.

• Journal of Broadcast Engineering
• /
• v.28 no.1
• /
• pp.124-131
• /
• 2023

Object detection techniques are increasingly used to obtain information on physical characteristics or situations of a specific area from remote images. The accuracy of object detection is decreased in remote sensing images with low resolution because the low resolution reduces the amount of detail that can be captured in an image. A single neural network is proposed to joint the super-resolution method and object detection method. The proposed method constructs a deep residual-based network to restore object features in low-resolution images. Moreover, the proposed method is used to improve the performance of object detection by jointing a single network with YOLOv5. The proposed method is experimentally tested using VEDAI data for low-resolution images. The results show that vehicle detection performance improved by 81.38% on mAP@0.5 for VISIBLE data.

• Journal of the Korean Society of Radiology
• /
• v.17 no.5
• /
• pp.693-699
• /
• 2023

Ultrasound is widely used in the medical field for non-destructive and non-invasive disease diagnosis. In order to improve the disease diagnosis accuracy of diagnostic medical images, improving spatial resolution is a very important factor. In this study, we aim to model the super resolution convolutional neural network (SRCNN) algorithm in ultrasound images and analyze its applicability in the medical diagnostic field. The study was conducted as an experimental study using Field II simulation and open source clinical liver hemangioma ultrasound imaging. The proposed SRCNN algorithm was modeled so that end-to-end learning can be applied from low resolution (LR) to high resolution. As a result of the simulation, we confirmed that the full width at half maximum in the phantom image using a Field II program was improved by 41.01% compared to LR when SRCNN was used. In addition, the peak to signal to noise ratio (PSNR) and structural similarity index (SSIM) evaluation results showed that SRCNN had the excellent value in both simulated and real liver hemangioma ultrasound images. In conclusion, the applicability of SRCNN to ultrasound images has been proven, and we expected that proposed algorithm can be used in various diagnostic medical fields.

• Journal of the Korean Society of Radiology
• /
• v.2 no.1
• /
• pp.23-29
• /
• 2008

The present study proposed a method that dissolves ultrasonographic images into multiple resolutions using wavelet conversion and a boundary detection filter and improves the quality of ultrasonographic images through boundary detection filtering. In order to reduce noises and strengthen edges, the proposed method adjusted selectivity coefficient by area step by step from a low resolution image obtained from wavelet converted images to a high resolution image and performed edge filtering in consideration of direction. Through this method, we generated a selective low pass filtering effect in areas except edges by decreasing the wavelet coefficient for pixels in spot areas, improved continuity by smoothing edges in the tangential direction, and enhanced contrast by thinning in the normal direction. Through an experiment, we compared the filtering method using a non linear anisotropic expansion model and the filtering method using wavelet contraction structure in single resolution.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.1C
• /
• pp.64-71
• /
• 2003

In this paper, we proposed an algorithm that shorten coding time maintaining image quality in JPEG2000, which is the standard, of still image compression. This method encodes only the bit plane selected as appropriate truncation point for output bitstream, obtained from estimation of frequency distribution for whole image. Wavelet characterized by multi-resolution has vertical, horizontal, and diagonal frequency components for each resolution. The frequency interrelation addressed above is maintained thorough whole level of resolution and represents the unique frequency characteristics for input image. Thus, using the frequency relation at highest level, we can pick the truncation point for the compression time decrease by estimating code bits at encoding each code block. Also, we reduced the encoding time using simply down sampling instead of low-pass filtering at low-levels which are not encoded in color component of lower energy than luminance component. From the proposed algorithm, we can reduce about 15~36% of encoding time maintaining PSNR 30$\pm$0.5㏈.