• ETRI Journal
• /
• v.41 no.6
• /
• pp.750-759
• /
• 2019

Presently, the extraction of hand-crafted features is still the dominant method in radar emitter recognition. To solve the complicated problems of selection and updation of empirical features, we present a novel automatic feature extraction structure based on deep learning. In particular, a convolutional neural network (CNN) is adopted to extract high-level abstract representations from the time-frequency images of emitter signals. Thus, the redundant process of designing discriminative features can be avoided. Furthermore, to address the performance degradation of a single platform, we propose the construction of an ensemble learning-based architecture for multi-platform fusion recognition. Experimental results indicate that the proposed algorithms are feasible and effective, and they outperform other typical feature extraction and fusion recognition methods in terms of accuracy. Moreover, the proposed structure could be extended to other prevalent ensemble learning alternatives.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.6 no.3
• /
• pp.157-167
• /
• 2011

Although vehicles become more intelligent for convenience and safety of drivers, traffic accidents are increased more and more. Especially, car-to-car single rear impacts in the urban area are increased rapidly because of driver inattention. To prevent rear impacts in the urban area, commercial automobile vendor applies the low-speed short-distance collision warning system. This paper presents low-speed short-distance collision warning algorithm for the city driving by using sensor fusion of laser sensor and ultrasonic sensor. An experiment using embedded microprocessor in the driving track was used to demonstrate the feasibility of the collision warning algorithm.

• Journal of the Korea Society of Computer and Information
• /
• v.24 no.5
• /
• pp.59-66
• /
• 2019

According to the KT telecommunication statistics, people stayed inside their houses on an average of 11.9 hours a day. As well as, according to NSC statistics in the united states, people regardless of age are injured for a variety of reasons in their houses. For purposes of this research, we have investigated an abnormal event detection algorithm to classify infrequently occurring behaviors as accidents, health emergencies, etc. in their daily lives. We propose a fusion method that combines three classification algorithms with vision pattern, audio pattern, and activity pattern to detect unusual user events. The vision pattern algorithm identifies people and objects based on video data collected through home CCTV. The audio and activity pattern algorithms classify user audio and activity behaviors using the data collected from built-in sensors on their smartphones in their houses. We evaluated the proposed individual pattern algorithm and fusion method based on multiple scenarios.

• Journal of Advanced Navigation Technology
• /
• v.27 no.6
• /
• pp.884-887
• /
• 2023

Recently, research has been actively conducted to overcome the limitations of high-priced single sensors and reduce costs through the convergence of low-cost multi-variable sensors. This paper estimates state variables through asynchronous Kalman filters constructed using CVXPY and uses Cvxpylayers to compare and learn state variables estimated from CVXPY with true value data to estimate filter parameters of low-cost sensors fusion.

• Journal of The Korean Association of Information Education
• /
• v.19 no.1
• /
• pp.139-148
• /
• 2015

The purpose of image fusion is to combine the relevant information from a set of images into a single image, where the resultant fused image will be more informative and complete than any of the input images. Image fusion techniques can improve the quality and increase the application of these data important applications of the fusion of images include medical imaging, remote sensing, and robotics. In this paper, we suggest a new method to generate a fusion image using the close relation of image features obtained through maximum entropy threshold and mutual information. This method represents a good image registration in case of using a blurring image than other image fusion methods.

• The Korean Journal of Physiology and Pharmacology
• /
• v.7 no.6
• /
• pp.375-379
• /
• 2003

$Ca^{2+}$ influx appears to be important for triggering myoblast fusion. It remains, however, unclear how $Ca^{2+}$ influx rises prior to myoblast fusion. Recently, several studies suggested that NMDA receptors may be involved in $Ca^{2+}$ mobilization of muscle, and that $Ca^{2+}$ influx is mediated by NMDA receptors in C2C12 myoblasts. Here, we report that other types of ionotropic glutamate receptors, non-NMDA receptors (AMPA and KA receptors), are also involved in $Ca^{2+}$ influx in myoblasts. To explore which subtypes of non-NMDA receptors are expressed in C2C12 myogenic cells, RT-PCR was performed, and the results revealed that KA receptor subunits were expressed in both myoblasts and myotubes. However, AMPA receptor was not detected in myoblasts but expressed in myotubes. Using a $Ca^{2+}$ imaging system, $Ca^{2+}$ influx mediated by these receptors was directly measured in a single myoblast cell. Intracellular $Ca^{2+}$ level was increased by KA, but not by AMPA. These results were consistent with RT-PCR data. In addition, KA-induced intracellular $Ca^{2+}$ increase was completely suppressed by treatment of nifedifine, a L-type $Ca^{2+}$ channel blocker. Furthermore, KA stimulated myoblast fusion in a dose-dependent manner. CNQX inhibited not only KA-induced myoblast fusion but also spontaneous myoblast fusion. Therefore, these results suggest that KA receptors are involved in intracellular $Ca^{2+}$ increase in myoblasts and then may play an important role in myoblast fusion.

• Journal of the Korean Society for Precision Engineering
• /
• v.27 no.2
• /
• pp.130-136
• /
• 2010

For stability analysis of the lumbar spine, the hypothesis presented is that the disc has stress sensors driving feedback mechanism, which could react to the imposed loads by adjusting the contraction of the muscles. Fusion in the motion segment of the lumbar spinal column is believed to alter the stability of the spinal column. To identify this effect finite element (FE) models combined with optimization technique was applied and quantify the role of each muscle and reaction forces in the spinal column with respect to the fusion level. The musculoskeletal FE model was consisted with detailed whole lumbar spine, pelvis, sacrum, coccyx and simplified trunk model. Vertebral body and pelvis were modeled as a rigid body and the rib cage was constructed with rigid truss element for the computational efficiency. Spinal fusion model was applied to L3-L4, L4-L5, L5-S1 (single level) and L3-L5 (two levels) segments. Muscle architecture with 46 local muscles was used as acting directions. Minimization of the nucleus pressure deviation and annulus fiber average axial stress deviation was selected for cost function. As a result, spinal fusion produced reaction changes at each motion segment as well as contribution of each muscle. Longissimus thoracis and psoas major muscle showed dramatic changes for the cases of L5-S1 and L3-L5 level fusion. Muscle force change at each muscle also generated relatively high nucleus pressure not only at the adjacent level but at another level, which can explain disc degeneration pattern observed in clinical study.

• Journal of Microbiology
• /
• v.45 no.1
• /
• pp.41-47
• /
• 2007

The Hantaan virus (HTNV) is an enveloped virus that is capable of inducing low pH-dependent cell fusion. We molecularly cloned the viral glycoprotein (GP) and nucleocapsid (NP) cDNA of HTNV and expressed them in Vero E6 cells under the control of a CMV promoter. The viral gene expression was assessed using an indirect immunofluorescence assay and immunoprecipitation. The transfected Vero E6 cells expressing GPs, but not those expressing NP, fused and formed a syncytium following exposure to a low pH. Monoclonal antibodies (MAbs) against envelope GPs inhibited cell fusion, whereas MAbs against NP did not. We also investigated the N-linked glycosylation of HTNV GPs and its role in cell fusion. The envelope GPs of HTNV are modified by N-linked glycosylation at five sites: four sites on G1 (N134, N235, N347, and N399) and one site on G2 (N928). Site-directed mutagenesis was used to construct eight GP gene mutants, including five single N-glycosylation site mutants and three double-site mutants, which were then expressed in Vero E6 cells. The oligosaccharide chain on residue N928 of G2 was found to be crucial for cell fusion after exposure to a low pH. These results suggest that G2 is likely to be the fusion protein of HTNV.

• Journal of Korean Neurosurgical Society
• /
• v.52 no.1
• /
• pp.27-31
• /
• 2012

Objective : Although hydroxyapatite (HA) spacer has been used for laminoplasty, there have been no reports on factors associated with fusion and on the effects of HA shape. Methods : During January 2004 and January 2010, 45 patients with compressive cervical myelopathy underwent midline-splitting open door laminoplasty with winged (33 cases) and wingless (12 cases) HAs by a single surgeon. Minimal and mean follow up times were 12 and 28.1 months, respectively. Japanese Orthopedic Association (JOA) score was used for clinical outcome measurement. Cervical X-rays were taken preoperatively, immediately post-operatively, and after 3, 6, and 12 months and computed tomography scans were performed preoperatively, immediately post-operatively and after 12 months. Cervical lordosis, canal dimension, fusion between lamina and HA, and affecting factors of fusion were analyzed. Results : All surgeries were performed on 142 levels, 99 in the winged and 43 in the wingless HA groups. JOA scores of the winged group changed from $10.4{\pm}2.94$ to $13.3{\pm}2.35$ and scores of the wingless group changed from $10.8{\pm}2.87$ to $13.8{\pm}3.05$. There was no significant difference on lordotic and canal dimensional change between two groups. Post-operative 12 month fusion rate between lamina and HA was significantly lower in the winged group (18.2 vs. 48.8% p=0.001). Multivariate analysis showed that ossification of the posterior longitudinal ligament, male gender, and wingless type HA were significantly associated with fusion. Conclusion : Clinical outcome was similar in patients receiving winged and wingless HA, but the wingless type was associated with a higher rate of fusion between HA and lamina at 12 months post-operatively.

• Journal of Broadcast Engineering
• /
• v.24 no.4
• /
• pp.580-591
• /
• 2019

This paper focuses on to improving object detection performance using the camera and LiDAR on autonomous vehicle platforms by fusing detected objects from individual sensors through a late fusion approach. In the case of object detection using camera sensor, YOLOv3 model was employed as a one-stage detection process. Furthermore, the distance estimation of the detected objects is based on the formulations of Perspective matrix. On the other hand, the object detection using LiDAR is based on K-means clustering method. The camera and LiDAR calibration was carried out by PnP-Ransac in order to calculate the rotation and translation matrix between two sensors. For Sensor fusion, intersection over union(IoU) on the image plane with respective to the distance and angle on world coordinate were estimated. Additionally, all the three attributes i.e; IoU, distance and angle were fused using logistic regression. The performance evaluation in the sensor fusion scenario has shown an effective 5% improvement in object detection performance compared to the usage of single sensor.