• Biomedical Engineering Letters
• /
• v.8 no.4
• /
• pp.373-382
• /
• 2018

Since epileptic seizure is unpredictable and paroxysmal, an automatic system for seizure detecting could be of great significance and assistance to patients and medical staff. In this paper, a novel method is proposed for multichannel patient-specific seizure detection applying the earth mover's distance (EMD) in scalp EEG. Firstly, the wavelet decomposition is executed to the original EEGs with five scales, the scale 3, 4 and 5 are selected and transformed into histograms and afterwards the distances between histograms in pairs are computed applying the earth mover's distance as effective features. Then, the EMD features are sent to the classifier based on the Bayesian linear discriminant analysis (BLDA) for classification, and an efficient postprocessing procedure is applied to improve the detection system precision, finally. To evaluate the performance of the proposed method, the CHB-MIT scalp EEG database with 958 h EEG recordings from 23 epileptic patients is used and a relatively satisfactory detection rate is achieved with the average sensitivity of 95.65% and false detection rate of 0.68/h. The good performance of this algorithm indicates the potential application for seizure monitoring in clinical practice.

• Journal of The Geomorphological Association of Korea
• /
• v.23 no.1
• /
• pp.103-115
• /
• 2016

At present, an acoustic Doppler current profiler (ADCP) is one of the most suitable tools for measurement of three dimensional flow characteristics in the river. The data resulting from this approach can be used for flow visualization and velocity mapping together with post-processing software tools. Among them, 'adcptools' is the latest one and provides more realistic velocity distribution in the cross-section since it uses velocity along the beam direction. In this study, a flow analysis was made using the 'adcptools' for the Amazon River and the Han River dataset. Discharge was recalculated and accuracy of discharge and velocity was evaluated. Streamwise velocity distribution and secondary flow pattern in cross-sections were visualized. Geo-referenced velocity distribution was also mapped. A summary with future prospect of 'adcptools' for studies on fluvial geomorphology is briefly given.

• Journal of Internet Computing and Services
• /
• v.22 no.3
• /
• pp.37-43
• /
• 2021

In this paper, we study how objects from front and back cameras can be recognized in real time in a mobile environment to segment regions of object pixels and synthesize them through image processing. To this work, we applied DeepLabV3 machine learning model to dual cameras provided by Apple's iOS. We also propose methods using Core Image and Core Graphics libraries from Apple for image synthesis and postprocessing. Furthermore, we improved CPU usage than previous works and compared the throughput rates and results of Depth and DeepLabV3. Finally, We also developed a camera application using these two methods.

• Journal of Korea Multimedia Society
• /
• v.24 no.8
• /
• pp.1114-1121
• /
• 2021

In this paper, we implement a recipe recommendation system using ingredient harmonization analysis based on recipe data. The proposed system receives an image of a food ingredient purchase receipt to recommend ingredients and recipes to the user. Moreover, it performs preprocessing of the receipt images and text extraction using the OCR algorithm. The proposed system can recommend recipes based on the combined data of ingredients. It collects recipe data to calculate the combination for each food ingredient and extracts the food ingredients of the collected recipe as training data. And then, it acquires vector data by learning with a natural language processing algorithm. Moreover, it can recommend recipes based on ingredients with high similarity. Also, the proposed system can recommend recipes using replaceable ingredients to improve the accuracy of the result through preprocessing and postprocessing. For our evaluation, we created a random input dataset to evaluate the proposed recipe recommendation system's performance and calculated the accuracy for each algorithm. As a result of performance evaluation, the accuracy of the Word2Vec algorithm was the highest.

• Journal of the Korean Society of Industry Convergence
• /
• v.26 no.4_1
• /
• pp.517-527
• /
• 2023

Many techniques have been proposed and investigated for microlens array manufacturing in three-dimensional (3D) structures. We present fabricating a microlens array using selective laser etching and a CO₂ laser. The femtosecond laser was employed to produce multiple micro-cracks that comprise the predesigned 3D structure. Subsequently, the wet etching process with a KOH solution was used to produce the primary microlens array structures. To polish the nonoptical surface to the optical surface, we performed reflow postprocessing using a CO₂ laser. We confirmed that the micro lens array can be manufactured in three primary shapes (cone, pyramid and hemisphere). Compared to our previous study, the processing time required for laser processing was reduced from approximately 1 hour to less than 30 seconds using the proposed processing method. Therefore, micro lens arrays can be manufactured using our processing method and can be applied to mass productionon large surface areas.

• Journal of Powder Materials
• /
• v.30 no.2
• /
• pp.116-122
• /
• 2023

Because magnets fabricated using Nd-Fe-B exhibit excellent magnetic properties, this novel material is used in various high-tech industries. However, because of the brittleness and low formability of Nd-Fe-B magnets, the design freedom of shapes for improving the performance is limited based on conventional tooling and postprocessing. Laser-powder bed fusion (L-PBF), the most famous additive manufacturing (AM) technique, has recently emerged as a novel process for producing geometrically complex shapes of Nd-Fe-B parts owing to its high precision and good spatial resolution. However, because of the repeated thermal shock applied to the materials during L-PBF, it is difficult to fabricate a dense Nd-Fe-B magnet. In this study, a high-density (>96%) Nd-Fe-B magnet is successfully fabricated by minimizing the thermal residual stress caused by substrate heating during L-PBF.

• ETRI Journal
• /
• v.45 no.5
• /
• pp.822-835
• /
• 2023

Recent semantic segmentation frameworks usually combine low-level and high-level context information to achieve improved performance. In addition, postlevel context information is also considered. In this study, we present a Context ReFinement Network (CRFNet) and its training method to improve the semantic predictions of segmentation models of the encoder-decoder structure. Our study is based on postprocessing, which directly considers the relationship between spatially neighboring pixels of a label map, such as Markov and conditional random fields. CRFNet comprises two modules: a refiner and a combiner that, respectively, refine the context information from the output features of the conventional semantic segmentation network model and combine the refined features with the intermediate features from the decoding process of the segmentation model to produce the final output. To train CRFNet to refine the semantic predictions more accurately, we proposed a sequential training scheme. Using various backbone networks (ENet, ERFNet, and HyperSeg), we extensively evaluated our model on three large-scale, real-world datasets to demonstrate the effectiveness of our approach.

• Journal of Cerebrovascular and Endovascular Neurosurgery
• /
• v.25 no.3
• /
• pp.253-259
• /
• 2023

Precise evaluation of the feeders, fistulous points, and draining veins plays a key role for successful embolization of intracranial dural arteriovenous fistulas (DAVF). Digital subtraction angiography (DSA) is a gold standard diagnostic tool to assess the exact angioarchitecture of DAVFs. With the advent of new image postprocessing techniques, we lately have been able to apply image fusion techniques with two different image sets obtained with flat panel detector rotational angiography. This new technique can provide additional and better pretherapeutic information of DAVFs over the conventional 2D and 3D angiographies. In addition, it can be used during the endovascular treatment to help the accurate and precise navigation of the microcatheter and microguidwire inside the vessels and identify the proper location of microcatheter in the targeted shunting pouch. In this study, we briefly review the process of an image fusion technique and introduce our clinical application for treating DAVFs, especially focused on the transvenous embolization.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.678-683
• /
• 2006

A novel and intuitive way of accessing applications of mobile devices is presented. The key idea is to use grip-pattern, which is naturally produced when a user tries to use the mobile device, as a clue to determine an application to be launched. To this end, a capacitive touch sensor system is carefully designed and installed underneath the housing of the mobile terminal to capture the image of the user's grip-pattern. The captured data is then recognized by a recognizer with dedicated preprocessing and postprocessing algorithms. The recognition test is performed to validate the feasibility of the proposed user interface system.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.43 no.12
• /
• pp.1048-1053
• /
• 2015

Ablative materials in a thermal protection system for atmospheric re-entry suffers from the most severe heat fluxes and temperatures, which induces surface recession in the thickness direction. In this paper, a 0.4MW arc-heated wind tunnel is operated to test for ablative materials, and a non-contact three-dimensional surface measuring system is used to evaluate the different surface characteristics of them. In particular, by postprocessing the three-dimensional image data, the surface roughness and recession of ablative materials can be calculated before and after the wind tunnel test. Moreover, the surface properties are analyzed quantitatively by comparing volume and mass losses of the test specimens.