• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.10
• /
• pp.3729-3749
• /
• 2021

At present, deep convolution network-based salient object detection (SOD) has achieved impressive performance. However, it is still a challenging problem to make full use of the multi-scale information of the extracted features and which appropriate feature fusion method is adopted to process feature mapping. In this paper, we propose a new adjacency auxiliary network (AANet) based on multi-scale feature fusion for SOD. Firstly, we design the parallel connection feature enhancement module (PFEM) for each layer of feature extraction, which improves the feature density by connecting different dilated convolution branches in parallel, and add channel attention flow to fully extract the context information of features. Then the adjacent layer features with close degree of abstraction but different characteristic properties are fused through the adjacent auxiliary module (AAM) to eliminate the ambiguity and noise of the features. Besides, in order to refine the features effectively to get more accurate object boundaries, we design adjacency decoder (AAM_D) based on adjacency auxiliary module (AAM), which concatenates the features of adjacent layers, extracts their spatial attention, and then combines them with the output of AAM. The outputs of AAM_D features with semantic information and spatial detail obtained from each feature are used as salient prediction maps for multi-level feature joint supervising. Experiment results on six benchmark SOD datasets demonstrate that the proposed method outperforms similar previous methods.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.9 no.6
• /
• pp.345-351
• /
• 2014

This paper presents pedestrian recognition to improve performance for vehicle safety system or surveillance system. Pedestrian detection method using HOG (Histograms of Oriented Gradients) has showed 90% recognition rate. But if someone takes a picture in the rain, the image may be distorted by rain streaks and recognition rate goes down by 62%. To solve this problem, we applied image decomposition method using MCA (Morphological Component Analysis). In this case, rain removal method improves recognition rate from 62% to 70%. However, it is difficult to apply conventional image decomposition method using MCA on vehicle safety system or surveillance system as conventional method is too slow for real-time system. To alleviate this issue, we propose a rain removal method by using low-pass filter and DCT (Discrete Cosine Transform). The DCT helps separate the image into rain components. The image is removed rain components by Butterworth filtering. Experimental results show that our method achieved 90% of recognition rate. In addition, the proposed method had accelerated processing time to 17.8ms which is acceptable for real-time system.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.6 no.4
• /
• pp.22-28
• /
• 2003

In the conventional weapon system, such as gun and small arms, it Is a general trend that for maximization of its performance and enhancement of its effectiveness, the firing control system(FCS) is developed and applied with the guns and small arms in the world. The FCS of the small arms for infantry man is composed of a few of sensors for acquisition of input data of FCS, such as range measurement, position sensing of weapon, temperature, etc., computer, displayer and power pack, and also the air burst munition is developed in parallel for the maximization of FCS's effectiveness. Since the flight time setting fuze for the air burst munition is adapted for next me, the measuring device of the muzzle velocity is needed to overcome the variation of muzzle velocity due to producing procedures and the differences of the using temperatures and so maintain the burst position accuracy This paper contained the technical information on the development of the measuring device of muzzle velocity, which designed in compact & light weight configuration with reliability and accuracy.

• Proceedings of the KIEE Conference
• /
• 1990.07a
• /
• pp.341-345
• /
• 1990

This paper presents the three section sliding mode control algorithm based on hysteresis current control add indirect field oriented control method, and applies it to the position control of induction motor. The three section sliding trajectories are defined in such a way that the system responds following a max acceleration line, then a max speed line, and finally a max deceleration line. This control scheme solves the problem of robustness loss during the reaching phase that occurs in conventional VSC strategy, and ensures the stable sliding mode and robustness enhancement throughout an entire response. Also, the PID controller operating in parallel is adopted to eliminate the sliding mode's collapse phenomenon near the origin caused by steady state chattering phenomenon Digital simulation results confirm that the dynamic performance of the system is insensitive to parameter variations and disturbances.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2005.11a
• /
• pp.159-160
• /
• 2005

The superconducting fault current limiter(SFCLs) provides the effect such as enhancement in the power system reliability due to limiting fault current in a few miliseconds. The Flux-lock type SFCL using the YBCO film among various type SFCLs consists of the primary and the secondary copper coils that are wound in parallel each other through the iron core. The operation can be controlled by adjusting the inductances and the winging directions of each the coil. We compared the current limiting performance on the additive and the subtractive polarity winding directions in case of an open-loop iron core. To analyze quench characteristics, we experimented various phase angle.

• Journal of the Korea Society of Computer and Information
• /
• v.4 no.4
• /
• pp.185-192
• /
• 1999

In this paper. for $\pi$/4 and $\pi$/8 PSK. we proposed to condition to obtain coding gain increasing states, by design encoder of analytical method with minimal complexity in limited bandwidth and power channels. In order to improve the bit error rate(BER), comparing Ungerboeck designed the TCM. we propose MTCM(Multiple trellis-coded modulation) with multiplicity(k＝2), by optimum encoder design. By design encoder of analytical method. the trellis encoder can be minimal complexity and the decoder be used Viterbi decoder(MLSE). When compared to the TCM and MTCM with AWGN channels. the condition of performance enhancement of the MTCM with multiplicity(k＝2) is the case of parallel transition in TCM systems. without alternating data transmission rate in bandwidth and power limited channels.

• Journal of the Semiconductor & Display Technology
• /
• v.15 no.2
• /
• pp.74-80
• /
• 2016

This study presents a thermal device specially designed for thermal nanoimprint lithography equipments, which requires the capability of rapid heating and cooling, high temperature uniformity and the material strength to endure high stamping pressure. The proposal to meet these requirements is a planar-type hot plate extensible to a large area, in which long circular cartridge heaters and heat pipes are installed inside in parallel. The heat pipes are connected to the outside water cooling chamber. A hot plate made of stainless steel is fabricated with a dimension $240mm{\times}240mm{\times}20mm$. Laboratory experiments are conducted to examine the thermal performance of the hot plate. The results illustrate that the employment of heat pipes leads to a notable enhancement of temperature uniformity in the device and provides an efficient heat delivery from the hot plate to outside. It is verified that the suggested hot plate could be a feasible thermal tool for thermal nanoimprint lithography, satisfying the major design requirements.

• Journal of KIISE:Software and Applications
• /
• v.41 no.7
• /
• pp.492-500
• /
• 2014

This paper presents the performance enhancement of automatic bilingual lexicon extraction by using refinement of pivot-context vectors under the standard pivot-based approach, which is very effective method for less-resource language pairs. In this paper, we gradually improve the performance through two different refinements of pivot-context vectors: One is to filter out unhelpful elements of the pivot-context vectors and to revise the values of the vectors through bidirectional translation probabilities estimated by Anymalign and another one is to remove non-noun elements from the original vectors. In this paper, experiments have been conducted on two different language pairs that are bi-directional Korean-Spanish and Korean-French, respectively. The experimental results have demonstrated that our method for high-frequency words shows at least 48.5% at the top 1 and up to 88.5% at the top 20 and for the low-frequency words at least 43.3% at the top 1 and up to 48.9% at the top 20.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.19 no.12
• /
• pp.1462-1469
• /
• 2008

This paper proposes the design method based on the Composite Right/Left-Handed(CRLH) structure to miniaturize and enhance the rejection performance in the stopband of the bandpass filter for a VHF band($169{\sim}211\;MHz$). For realization, we used the 0-th order resonance point of the CRLH and the simple theory of Inverter as coupling. The proposed technique is validated by the performance predictions and experiments, (Insertion loss <2 dB, $S_{11}$< -15 dB, suppression of up to the 3rd harmonic) and it is found out that the suggested method enables the size reduction of around 60 % from the conventional filters such as the parallel edge coupled type.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.24 no.5
• /
• pp.744-752
• /
• 2000

The influence of arrangement and length of discrete ribs on heat/mass transfer and friction loss is investigated. Mass transfer experiments are conducted to obtain the detailed local heat/mass transfer information on the ribbed wall. The aspect ratio (width/height) of the duct is 2.04 and the rib height is one tenth of the duct height, such that the ratio of the rib height to hydraulic diameter is 0.0743. The ratio of rib-to-rib distance to rib height is 10. The discrete ribs were made by dividing each continuous rib into 2, 3 or 5 pieces and attached periodically to the top and the bottom walls of the duct with a parallel orientation The combined effects of rib angle and length of the discrete ribs on heat/mass transfer ae considered for the rib angles $({\alpha})\;of\;90^{\circ}\;and\;45^{\circ}$. As the number of the discrete ribs increases, the uniformity of the heat/mass transfer distributions increases. For $(\alpha})=90^{\circ}$, the heat/mass transfer enhancement with the discrete ribs is remarkable, while the heat/mass transfer performances are slightly higher than that of the transverse continuous ribs due to the accompanied high friction loss penalty. For $(\alpha})=90^{\circ}$, the average heat/mass transfer coefficients and the heat/mass transfer performances decrease slightly with the discrete ribs compared to the case of the angled continuous ribs.