• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.7
• /
• pp.838-846
• /
• 2004

In the case where both a camera and a target are moving at the same time, the image background is successively changed, and the overlap with other moving objects is apt to be generated. The snake algorithms have been variously used in tracking the object, but it is difficult to be applied in the excessive overlap with other objects and the large bias between the snake and the target. To solve this problem, this paper presents an extended snake model. It includes an additional energy function which considers the temporal variation rate of the snake's area and a SSD algorithm which generates the template adaptive to the snake detected in the previous frame. The new energy function prevents the snake from over-shrinking or stretching and the SSD algorithm with adaptively changing template allows the prediction of the target's position in the next frame. The experimental results have shown that the proposed algorithm successfully tracks the target even when the target is temporarily occluded by other objects.

• Korean Journal of Remote Sensing
• /
• v.15 no.2
• /
• pp.119-130
• /
• 1999

At present, many ground stations all over the world are using NORAD orbit element data in order to track and communicate with Earth orbiting satellites. The North American Aerospace Defense Command (NORAD) observes thousands of Earth orbiting objects on daily basis and provides their orbital information via internet. The orbital data provided by NORAD, which is also called two line element (TLE) sets, allows ground stations to predict the time-varying positions of satellites accurately enough to communicate with the satellites. In order to complete the mission of a high resolution remote sensing satellite which requires very high positional determination and control accuracy, however, a mission control and tracking ground station is dedicated for the observation and positional determination of the satellite rather than using NORAD orbital sets. In the case of KITSAT-3, NORAD orbital elements are currently used for image acquisition planning and for the processing of acquired images due to the absence of a dedicated KITSAT-3 tracking ground system. In this paper, we tested and analyzed the accuracy of NORAD orbital elements and the appropriate prediction model to determine how accurately a satellite acquisites an image of the location of interest and how accurately a ground processing system can generate the catalog of the images.

• The Journal of the Acoustical Society of Korea
• /
• v.14 no.2E
• /
• pp.73-86
• /
• 1995

The transducer section of the forward-looking ultrasound imaging catheter (FLUIC) consists of a circular piezoelectric element as a vibrator and a conical acoustic mirror as a perfect reflector. A small diameter piezoelectric transducer element is mounted on the side of a catheter's rotating shaft. The unique design of FLUIC provides the capability to form a two-dimensional image of a cross-section of vessel in front of the catheter, which is lacking in the present generation of intravascular ultrasound (IVUS) transducers, as well as a conventional side view image. The mirror configuration for the transducer section of the FLUIC is designed using an approximated ray tracing techniques. The diffraction transfer function approach [1] developed for the field prediction from primary sources is generalized and extended to predict the secondary diffraction characterstics from an acoustic mirror. The extended model is verified by simulation and experiment through a simple plane reflector and employed to analyzed the field characteristics of a FLUIC.

• Journal of Digital Contents Society
• /
• v.19 no.1
• /
• pp.103-112
• /
• 2018

In this paper, we propose a method of recognizing emotions using PPG sensor which measures blood flow according to emotion. From the existing PPG signal, we use a method of determining positive emotions and negative emotions in the frequency domain through PSD (Power Spectrum Density). Based on James R. Russell's two-dimensional prototype model, we classify emotions as joy, sadness, irritability, and calmness and examine their association with the magnitude of energy in the frequency domain. It is significant that this study used the same PPG sensor used in wearable devices to measure the top four kinds of emotions in the frequency domain through image experiments. Through the questionnaire, the accuracy, the immersion level according to the individual, the emotional change, and the biofeedback for the image were collected. The proposed method is expected to be various development such as commercial application service using PPG and mobile application prediction service by merging with context information of existing smart phone.

• Smart Media Journal
• /
• v.10 no.1
• /
• pp.32-38
• /
• 2021

Image classification through deep learning on the image from photographs has been a very active research field for the past several years. In this paper, we propose a method of automatically discriminating stone images from domestic source through deep learning, which is to use Python's hash library to scan 300×300 pixel photo images of granites such as Hwangdeungseok, Goheungseok, and Pocheonseok, performing data preprocessing to create learning images by examining duplicate images for each stone, removing duplicate images with the same hash value as a result of the inspection, and deep learning by stone. In addition, to utilize VGGNet, the size of the images for each stone is resized to 224×224 pixels, learned in VGG16 where the ratio of training and verification data for learning is 80% versus 20%. After training of deep learning, the loss function graph and the accuracy graph were generated, and the prediction results of the deep learning model were output for the three kinds of stone images.

• Geophysics and Geophysical Exploration
• /
• v.23 no.3
• /
• pp.168-177
• /
• 2020

Seabed multiple reflections (seabed multiples) are the main cause of misinterpretations of primary reflections in both shot gathers and stack sections. Accordingly, seabed multiples need to be suppressed throughout data processing. Conventional model-driven methods, such as prediction-error deconvolution, Radon filtering, and data-driven methods, such as the surface-related multiple elimination technique, have been used to attenuate multiple reflections. However, the vast majority of processing workflows require time-consuming steps when testing and selecting the processing parameters in addition to computational power and skilled data-processing techniques. To attenuate seabed multiples in seismic reflection data, input gathers with seabed multiples and label gathers without seabed multiples were generated via numerical modeling using the Marmousi2 velocity structure. The training data consisted of normal-moveout-corrected common midpoint gathers fed into a U-Net neural network. The well-trained model was found to effectively attenuate the seabed multiples according to the image similarity between the prediction result and the target data, and demonstrated good applicability to field data.

• Smart Media Journal
• /
• v.11 no.10
• /
• pp.65-75
• /
• 2022

Deep learning is used as a creative tool that could overcome the limitations of existing analysis models and generate various types of results such as text, image, and music. In this paper, we propose a method necessary to preprocess audio data using the Niko's MIDI Pack sound source file as a data set and to generate music using Bi-LSTM. Based on the generated root note, the hidden layers are composed of multi-layers to create a new note suitable for the musical composition, and an attention mechanism is applied to the output gate of the decoder to apply the weight of the factors that affect the data input from the encoder. Setting variables such as loss function and optimization method are applied as parameters for improving the LSTM model. The proposed model is a multi-channel Bi-LSTM with attention that applies notes pitch generated from separating treble clef and bass clef, length of notes, rests, length of rests, and chords to improve the efficiency and prediction of MIDI deep learning process. The results of the learning generate a sound that matches the development of music scale distinct from noise, and we are aiming to contribute to generating a harmonistic stable music.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.36 no.5
• /
• pp.339-346
• /
• 2023

The uncertainties of microstructural features affect the properties of materials. Numerous pores that are randomly distributed in materials make it difficult to predict the properties of the materials. The distribution of pores in cementitious materials has a great influence on their mechanical properties. Existing studies focus on analyzing the statistical relationship between pore distribution and material responses, and the correlation between them is not yet fully determined. In this study, the mechanical response of cementitious materials is predicted through an image-based data approach using a convolutional neural network (CNN), and the correlation between pore distribution and material response is analyzed. The dataset for machine learning consists of high-resolution micro-CT images and the properties (tensile strength) of cementitious materials. The microstructures are characterized, and the mechanical properties are evaluated through 2D direct tension simulations using the phase-field fracture model. The attributes of input images are analyzed to identify the spot with the greatest influence on the prediction of material response through CNN. The correlation between pore distribution characteristics and material response is analyzed by comparing the active regions during the CNN process and the pore distribution.

• Korean Journal of Radiology
• /
• v.22 no.3
• /
• pp.425-434
• /
• 2021

Objective: To investigate the potential value of ¹⁸F-fluorodeoxyglucose (FDG) PET/CT in predicting the survival of patients with primary tracheal malignant tumors. Materials and Methods: An analysis of FDG PET/CT findings in 37 primary tracheal malignant tumor patients with a median follow-up period of 43.2 months (range, 10.8-143.2 months) was performed. Cox proportional hazards regression analyses were used to assess the associations between quantitative ¹⁸F-FDG PET/CT parameters, other clinic-pathological factors, and overall survival (OS). A risk prognosis model was established according to the independent prognostic factors identified on multivariate analysis. A survival curve determined by the Kaplan-Meier method was used to assess whether the prognosis prediction model could effectively stratify patients with different risks factors. Results: The median survival time of the 37 patients with tracheal tumors was 38.0 months, with a 95% confidence interval of 10.8 to 65.2 months. The 3-year, 5-year and 10-year survival rate were 54.1%, 43.2%, and 16.2%, respectively. The metabolic tumor volume (MTV), total lesion glycolysis (TLG), maximum standardized uptake value, age, pathological type, extension categories, and lymph node stage were included in multivariate analyses. Multivariate analysis showed MTV (p = 0.011), TLG (p = 0.020), pathological type (p = 0.037), and extension categories (p = 0.038) were independent prognostic factors for OS. Additionally, assessment of the survival curve using the Kaplan-Meier method showed that our prognosis prediction model can effectively stratify patients with different risks factors (p < 0.001). Conclusion: This study shows that ¹⁸F-FDG PET/CT can predict the survival of patients with primary tracheal malignant tumors. Patients with an MTV > 5.19, a TLG > 16.94 on PET/CT scans, squamous cell carcinoma, and non-E1 were more likely to have a reduced OS.

• Information Systems Review
• /
• v.22 no.4
• /
• pp.21-39
• /
• 2020

Recently, deep learning has shown high performance in various applications such as pattern analysis and image classification. Especially known as a difficult task in the field of machine learning research, stock market forecasting is an area where the effectiveness of deep learning techniques is being verified by many researchers. This study proposed a deep learning Convolutional Neural Network (CNN) model to predict the direction of stock prices. We then used the feature selection method to improve the performance of the model. We compared the performance of machine learning classifiers against CNN. The classifiers used in this study are as follows: Logistic Regression, Decision Tree, Neural Network, Support Vector Machine, Adaboost, Bagging, and Random Forest. The results of this study confirmed that the CNN showed higher performancecompared with other classifiers in the case of feature selection. The results show that the CNN model effectively predicted the stock price direction by analyzing the embedded values of the financial data