• Journal of Korean Ophthalmic Optics Society
• /
• v.10 no.4
• /
• pp.419-428
• /
• 2005

The aim of this study was to provide data for the relief of asthenopia during binocular vision by determining the characteristics of ocular function in adults. A total of 260 subjects were between the age of 19-35years. We measured individually the refractive error correction, pupillary distance, optical center distance, phoria, convergence, accommodation and the AC/A as well as the asthenopia during binocular vision using a questionnaire. After analysis of factors affecting asthenopia, we also examined the reductive effect of asthenopia in subjects who had asthenopia using prism. To determine the factors affecting asthenopia during binocular vision, statistic analyses were carried out the multivariate Logistic regression model. The results of this study were as follow. The asthenopia during binocular vision was found 26.9% of subjects. Multivariate logistic regression model was used to determine factors affecting binocular vision of myopia. When the accommodation and convergence were low compared to being high, when subjects had esophoria or there was more exophoria, and when AC/A was lower than the standard, the rate of asthenopia was higher. Therefore the accommodation, convergence and AC/A could be predictive factors for asthenopia. We used prism for subjects who had asthenopia during binocular vision, the results showed that the symptom of asthenopia was eased up to 74.3%.

• Journal of Sensor Science and Technology
• /
• v.12 no.6
• /
• pp.231-240
• /
• 2003

The problem of power consumption and the limitation of a chip area should be considered when the pixel number of the edge detection circuit increases to fabricate a vision chip for edge detection with high resolution. The numeric increment of the unit circuit causes power consumption to increase and require a larger chip area. An increment of power consumption and a limitation of chip area with several ten milli-meters square supplied by the CMOS foundry company restrict the pixel numbers of the edge detection circuit. In this paper, we proposed a electronic switch to minimize the power consumption owing to the numeric increment of the edge detection circuit to realize a vision chip for edge detection with high resolution. We also applied a method by which photodetector and edge detection circuit are separated to implement a vision chip with a higher resolution. The photodetector circuit with $128{\times}128$ pixels uses a common edge detection circuit with $1{\times}128$ pixels so that resolution was improved at the same chip area. The chip size is $4mm{\times}4mm$ and the power consumption was confirmed to be about 20mW using SPICE.

• Korean Journal of Ophthalmology
• /
• v.32 no.6
• /
• pp.451-458
• /
• 2018

Purpose: To investigate the efficacy, safety, and anatomical outcomes associated with intravitreal anti-vascular endothelial growth factor (VEGF) treatment of retinopathy of prematurity (ROP). Methods: We performed a retrospective review of intravitreal anti-VEGF (bevacizumab or ranibizumab) treatment of 153 eyes (83 infants) diagnosed with ROP at two tertiary hospitals from June 2011 to January 2017. The primary outcome was the rate of recurrence requiring additional treatment; secondary outcomes included incidence of major complications and final refractive error. Results: A total of 101 eyes were treated with bevacizumab, and 52 with ranibizumab. The bevacizumab and ranibizumab groups were characterized by mean birthweights of $941.8{\pm}296.1$ and $1,257.7{\pm}514.5g$, gestational ages at birth of $26.9{\pm}1.9$ and $28.1{\pm}3.2$ weeks, and postmenstrual ages at treatment of $40.4{\pm}2.4$ and $39.2{\pm}2.3$ weeks, respectively. The two groups differed significantly in birthweights and gestational ages at birth, but not in postmenstrual ages at treatment. The mean follow-up duration was $30.9{\pm}18.4$ months for the bevacizumab group, and $13.9{\pm}12.5$ months for ranibizumab. More cases were classified as zone 1 ROP in the ranibizumab group (44.2% vs. 11.9%, p < 0.001). Major surgical interventions included scleral encircling and vitrectomy (one and two eyes, respectively, both in the bevacizumab group). Retinal detachment was noted in one eye treated with bevacizumab. There was no significant difference in the most recent spherical equivalence for the two groups ($+0.10{\pm}3.66$ and $+0.22{\pm}3.00$ diopters for bevacizumab and ranibizumab, respectively). Univariable analysis revealed that only ROP stage influenced the occurrence of major complications (odds ratio, 9.046; p = 0.012). Conclusions: Intravitreal anti-VEGF treatment of ROP with both bevacizumab and ranibizumab achieved stable retinal vascularization with a low rate of complications and recurrence. Ranibizumab achieved similar anatomical outcomes as bevacizumab, without additional risk for major complications.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.7
• /
• pp.11-20
• /
• 2024

In this paper, we propose the BAG (Bit-width Aware Generator) and the Intermediate Layer Knowledge Distillation using Channel-wise Attention to reduce the knowledge gap between a quantized network, a full-precision network, and a generator in GDFQ (Generative Data-Free Quantization). Since the generator in GDFQ is only trained by the feedback from the full-precision network, the gap resulting in decreased capability due to low bit-width of the quantized network has no effect on training the generator. To alleviate this problem, BAG is quantized with same bit-width of the quantized network, and it can generate synthetic images, which are effectively used for training the quantized network. Typically, the knowledge gap between the quantized network and the full-precision network is also important. To resolve this, we compute channel-wise attention of outputs of convolutional layers, and minimize the loss function as the distance of them. As the result, the quantized network can learn which channels to focus on more from mimicking the full-precision network. To prove the efficiency of proposed methods, we quantize the network trained on CIFAR-100 with 3 bit-width weights and activations, and train it and the generator with our method. As the result, we achieve 56.14% Top-1 Accuracy and increase 3.4% higher accuracy compared to our baseline AdaDFQ.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.17 no.4
• /
• pp.323-335
• /
• 1992

In this paper, an adaptive cosine transform coding scheme which incorporate human visual properties into the coding scheme is investigated. Human vision is relatively sensitive to mid-frequency band, and insensitive to very low and very high frequency band. These property was mathematically modelled with MTF(Modulation Transfer Function) through many psychovisual experiment. DCT transforms energy in spatial domain into frequency domain, so can exploit the MTF very efficiently. Another well-known visual characteristics is spatial masking effect that visibility of noise is less in regions of high activity than in regions of low activity. Proposed coding scheme imploys quantization matrix which represent the properties of these spatial frequency response of human vision, and adaptively quality of an image. To compute the activity index of an image block, simple operation is performed in spatial domain, and according to activity index. block of low activity region is more exactly quantized relatively than that of high activity region. Results showed that, at low bit rate, the subjective quality of the reconstructed images by proposed coding scheme is acceptible than that of coding scheme without HVS properties.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.7
• /
• pp.962-970
• /
• 2021

The cause of the majority of vehicle accidents is a safety issue due to the driver's inattention, such as drowsy driving. A forward collision warning system (FCWS) can significantly reduce the number and severity of accidents by detecting the risk of collision with vehicles in front and providing an advanced warning signal to the driver. This paper describes a low power embedded system based FCWS for safety. The algorithm computes time to collision (TTC) through detection, tracking, distance calculation for the vehicle ahead and current vehicle speed information with a single camera. Additionally, in order to operate in real time even in a low-performance embedded system, an optimization technique in the program with high and low levels will be introduced. The system has been tested through the driving video of the vehicle in the embedded system. As a result of using the optimization technique, the execution time was about 170 times faster than that when using the previous non-optimized process.

• Journal of the Institute of Convergence Signal Processing
• /
• v.25 no.1
• /
• pp.33-38
• /
• 2024

The importance of connected technology that connects products or systems is increasing. Connected technology is a concept that can communicate with surrounding objects and connect to each other through a network to operate as one system. In particular, it can be implemented using wireless communication, and various conditions are required depending on the application system, such as communication distance and speed. In this study, we analyzed trends in communication technology for the implementation of connectivity between mobility devices such as self-driving cars, drones, UAVs, and shared mobility devices. The communication distance, speed, wired and wireless status, etc. of the latest communication methods currently commercialized or under development were investigated, with a particular focus on low-power operation. We identified the element technologies needed to build a low-power long-distance communication (LPWAN) system, and initially developed a plan for constructing a connected drone. The analysis results showed that it was possible to implement a connected system using the LoRa system, and an example configuration method was presented.

• Journal of the Korean Society for Aviation and Aeronautics
• /
• v.22 no.1
• /
• pp.145-151
• /
• 2014

Not only in Korea, but also in other countries, most helicopter accidents occur at low altitude. Analysis on accidents data collected both in Korea and other countries has brought two conclusions. Firstly, helicopter operations during the night hours carries a high risk. Secondly, the most common cause of night hour operations is loss of control due to the flight illusion. As an operation relying on a night vision in particular has a very high accident hazard, accompanying instruments such as NVG are ought to be provided. Hence, a thorough preparation and inspection on missions for night forest fire extinguishing should be required and perfect guidelines/road maps and enough training programs for the operation should be provided before the engagement in missions.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2003.05a
• /
• pp.409-418
• /
• 2003

The dies & molds are a very economical production tool and a high value-added product because of its mass production capability compared to other production methods. Due to the very stiff international competitiveness, the industries meet many obstacles. At this moment, measured the status, the international trends and try to find the vision and strategy to promotion of the industry. In Korea, large portion of dies and molds are still low value-added and make large number of tools. In order to advance this industry to the high-tech area and gain competitiveness in the golbal marketplaces, effective means of resource investment and strategy should be properly provided.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.64 no.3
• /
• pp.187-192
• /
• 2015

Detecting human in images is a challenging task owing to their variable appearance and the wide range of poses the they can adopt. The first need is a robust feature set that allows the human form to be discriminated cleanly, even in cluttered background under difficult illumination. A large number of vision application rely on matching keypoints across images. These days, the deployment of vision algorithms on smart phones and embedded device with low memory and computation complexity has even upped the ante: the goal is to make descriptors faster compute, more compact while remaining robust scale, rotation and noise. In this paper we focus on improving the speed of pedestrian(walking person) detection using Histogram of Oriented Gradient(HOG) descriptors provide excellent performance and tracking using kalman filter.