• The Journal of the Acoustical Society of Korea
• /
• v.42 no.4
• /
• pp.305-312
• /
• 2023

Physics-Informed Neural Network (PINN) is used to invert bubble size distributions from attenuation losses. By considering a linear system for the bubble population inversion, Adaptive Learned Iterative Shrinkage Thresholding Algorithm (Ada-LISTA), which has been solved linear systems in image processing, is used as a neural network architecture in PINN. Furthermore, a regularization based on the linear system is added to a loss function of PINN and it makes a PINN have better generalization by a solution satisfying the bubble physics. To evaluate an uncertainty of bubble estimation, deep ensemble is adopted. 20 Ada-LISTAs with different initial values are trained using the same training dataset. During test with attenuation losses different from those in the training dataset, the bubble size distribution and corresponding uncertainty are indicated by average and variance of 20 estimations, respectively. Deep ensemble Ada-LISTA demonstrate superior performance in inverting bubble size distributions than the conventional convex optimization solver of CVX.

• Journal of the Korean Society of Radiology
• /
• v.17 no.2
• /
• pp.207-213
• /
• 2023

Computed tomography (CT) images are used as the basis for proton Bragg peak position estimation and treatment plan simulation. During the Hounsfield Unit (HU) based proton stopping power ratio (SPR) estimation, small differences in the patient's density and elemental composition lead to uncertainty in the Bragg peak positions along the path of the proton beam. In this study, we investigated the potential of dual-energy computed tomography image-based proton SPRs prediction accuracy to reduce the uncertainty of Bragg peak position prediction. Single- and dual-energy images of an electron density phantom (CIRS Model 062M electron density phantom, CIRS Inc., Norfolk, VA, USA) were acquired using a computed tomography system (Somatom Definition AS, Siemens Health Care, Forchheim, Germany) to estimate the SPRs of the proton beam. To validate the method, it was compared to the SPRs estimated from standard data provided by the National Institute of Standards and Technology (NIST). The results show that the dual-energy image-based method has the potential to improve accuracy in predicting the SPRs of proton beams, and it is expected that further improvements in predicting the position of the proton's Bragg peak will be possible if a wider variety of substitutes with different densities and elemental compositions of the human body are used to predict the SPRs.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.12
• /
• pp.2924-2932
• /
• 2015

Image binarization is a process to divide the image into objects and backgrounds, widely applied to the fields of image analysis and its recognition. In the existing method of binarization, there is some uncertainty when there is insufficient brightness gap between objects and backgrounds in setting threshold. The method of fuzzy binarization has improved the features of objects efficiently. However, since this method sets ${\alpha}$-cut value statically, there remain some problems that important features of objects can be lost during binarization. Therefore, in this paper, we propose a binarization method which does not set ${\alpha}$-cut value statically. The proposed method uses fuzzy membership functions calculated by thresholds of mean, iterative, and Otsu binarization. Experiment results show the proposed method binaries various images with less loss than the existing methods.

• Ecology and Resilient Infrastructure
• /
• v.8 no.2
• /
• pp.88-96
• /
• 2021

In surface image velocimetry, a wide area of a river is photographed at an angle to measure its velocity, inevitably causing image distortion. Although a distorted image can be corrected into an orthogonal image by using 2D projective coordinate transformation and considering reference points on the same plane as the water surface, this method is limited by the uncertainty of changes in the water level in the event of a flood. Therefore, in this study, we developed a tilt image correction technique that corrects distortions in oblique images without resetting the reference points while coping with changes in the water level using the geometric relationship between the coordinates of the reference points set at a high position the camera, and the vertical distance between the water surface and the camera. Furthermore, we developed a distortion correction method to verify the corrected image, wherein we conducted a full-scale river experiment to verify the reference point transformation equation and measure the surface velocity. Based on the verification results, the proposed tilt image correction method was found to be over 97% accurate, whereas the experiment result of the surface velocity differed by approximately 4% as compared to the results calculated using the proposed method, thereby indicating high accuracy. Application of the proposed method to an image-based fixed automatic discharge measurement system can improve the accuracy of discharge measurement in the event of a flood when the water level changes rapidly.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.2
• /
• pp.160-167
• /
• 2016

This paper reports a localization method of a mobile robot using ceiling image. The ceiling has landmarks which are not distinguishablefrom one another. The location of every landmark in a map is given a priori while correspondence is not given between a detected landmark and a landmark in the map. Only the initial pose of the robot relative to the landmarks is given. The method uses particle filter approach for localization. Along with estimating robot pose, the method also associates a landmark in the map to a landmark detected from the ceiling image. The method is tested in an indoor environment which has circular landmarks on the ceiling. The test verifies the feasibility of the method in an environment where range data to walls or to beacons are not available or severely corrupted with noise. This method is useful for localization in a warehouse where measurement by Laser range finder and range data to beacons of RF or ultrasonic signal have large uncertainty.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.5 no.3
• /
• pp.117-127
• /
• 1995

In this paper, we suggest a motion analysis method using ART-I1 competitive learning neural network and fuzzy reasoning by matching the same objects through the consecutive image sequence. we use the size and mean intensity of the region obtained from image segmentation for the region matching by the region and use a ART-I1 competitive learning neural network wh~ch has a learning ability to reflect the topology of the input patterns in order to select characteristic points to describe the shape of a region. Motion vectors for each regions are obtained by matching selected characteristic points. However, the two dimensional image, the projection of the the three dimensional real world, produces fuzziness in motion analysis due to its incompleteness by nature and the error from image segmentation used for extracting information about objects. Therefore, the belief degrees for each regions are calculated using fuzzy reasoning to l-nanipulate uncertainty in motion estimation.

• The Journal of Korea Robotics Society
• /
• v.13 no.1
• /
• pp.26-30
• /
• 2018

Planetary global localization is necessary for long-range rover missions in which communication with command center operator is throttled due to the long distance. There has been number of researches that address this problem by exploiting and matching rover surroundings with global digital elevation maps (DEM). Using conventional methods for matching, however, is challenging due to artifacts in both DEM rendered images, and/or rover 2D images caused by DEM low resolution, rover image illumination variations and small terrain features. In this work, we use train CNN discriminator to match rover 2D image with DEM rendered images using conditional Generative Adversarial Network architecture (cGAN). We then use this discriminator to search an uncertainty bound given by visual odometry (VO) error bound to estimate rover optimal location and orientation. We demonstrate our network capability to learn to translate rover image into DEM simulated image and match them using Devon Island dataset. The experimental results show that our proposed approach achieves ~74% mean average precision.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.6
• /
• pp.1269-1274
• /
• 2014

In computer vision research, binarization procedure is one of the most frequently used tools to discriminate target objects from background in grey level binary image. Fuzzy binarization is a reliable technique in environment with high uncertainty such as medical image analysis by setting the threshold as the average of minimum and maximum brightness with triangle type fuzzy membership function. However, this technique is also known as contrast sensitive method thus its discrimination power is not so great when the image has low contrast difference between objects and backgrounds and suffer from information loss as a result. Thus, in this paper, we propose a fuzzy binarization using ART2 algorithm to handle such low contrast image analysis. Proposed ART2 algorithm is applied to determine the medium point of membership function in the fuzzy binarization paradigm. The proposed methods shows low information loss rate in our experiment.

• The Journal of Korean Society for Radiation Therapy
• /
• v.31 no.2
• /
• pp.25-31
• /
• 2019

Purpose: Metals induce metal artifact during CT-image for therapy planning, and it occurs images distortion, which affects the volumetric measurement and radiation calculation. In the case of using megavoltage computed tomography(MVCT), the volume of metals can be measured as similar to true volume due to minimal metal artifact outcome. In this study, radiation assessment was conducted by comparing teeth volume from images of kVCT and MVCT of head and neck cancer patients, then assigning to kVCT image to calculate radiation after obtaining the similar volume of true teeth volume from MVCT. Also, formal IR image was able to verify the accuracy of radiation calculation. Material and method: 5 head and neck cancer patients who had intensity-modulated radiation therapy from Radixact^® Series were of the subject in this study. Calculations of radiation when constraining true teeth volume out of kVCT image(A-CT) and when designated specific HU after teeth assigned using MVCT image were compared with formal IR image. Treatment planning was devised at the same constraints and mean dose was measured at the radiation assess points. The points were anterior of the teeth, between PTV and the teeth, the interior of PTV near the teeth, and the teeth where 5cm distance from PTV. Result: A difference of metals volume from kVCT and MVCT image was mean 3.49±2.61cc, maximum 7.43cc. PTV was limited to where the internal teeth were fully contained. The results of PTV dose evaluation showed that the average CI value of the kVCT treatment planning without the artifact correction was 0.86, and the average CI value of the kVCT with the artifact correction using MVCT image was 0.9. Conclusion: When the Treatment Planning was made without correction of metal artifacts, the dose of PTV was underestimated, indicating that dose uncertainty occurred. When the computerized treatment plan was made without correction of metal artifacts, the dose of PTV was underestimated, indicating that dose uncertainty occurred.

• Journal of the Korean Society of Costume
• /
• v.50
• /
• pp.51-68
• /
• 2000

This study is intended to examine the common features of costume images in science-fiction (SF) movies that deal with current socio-cultural situations by examining their themes and tones about the future it can be generally concluded that costume images of SF movies are divided into two patterns : one inheriting traditional styles constructed on linear progress and the other based on dismantiling the tradition. this analysis is made through the research of actual cinematic contexts on the common features of multiple styles shaping the two patterns of costume images. The results can be summarized as the following: The former is related with the future built up on the basis of belief in reasonal progress rooted in the Enlightenment reasonable plan for ideal social order and strong faith in uniformity. So It shows functional uniformity disregarding wasteful competitiveness in consumption and luxuriousness and clothing that has the aesthetic value of purity without emphasizing human body or sensuality are presented. On the other hand SF movies which show the uncertain costume image as the meaning of dismantling of tradition take up a rather critical view of assumption that society can move toward utopian future as it searches future images in the notion of hetero-topia by emphasizing pluralism consequently as for clothing diversity and uncertainty in post-modern style are presented destroying modernistic dichotomy and the assumption of Utopian clothing made in the notion of modern progressivism.