• Journal of the Korean Society of Radiology
• /
• v.14 no.7
• /
• pp.975-980
• /
• 2020

Cervical neural foraminal stenosis is a very common spinal disease that affects a relatively large number of people of all ages. However, since imaging methods that quantitatively provide neural foraminal stenosis are lacking, this study attempts to present quantitative measurement results by reconstructing 3D computed tomography images. Using a 3D reconstruction software, the surrounding bones were removed, including the spinous process, transverse process, and lamina of the cervical spine so that the neural foramen were well observed. Using Image J, a region of interest including the neural foramen area of the 3D image was set, and the number of pixels of the neural foramen area was measured. The neural foramen area was calculated by multiplying the number of measured pixels by the pixel size. In order to measure the widest area of the neural foramen, it was measured between 40-50 degrees in the opposite direction and 15-20 degrees toward the head. The measured cervical neural foramen area showed consistent measurement values. The largest measured area of the right neural foramen C5-6 was 12.21 ㎟, and after 2 years, the area was measured to be 9.95 ㎟, indicating that 18% stenosis had progressed. Since 3D reconstruction using axial CT scan images, no additional radiation exposure is required, and the area of stenosis can be objectively presented. In addition, it is good to explain to patients with neural stenosis while viewing 3D images, and it is considered a good method to be used in the evaluation of the progression of stenosis and post-operative evaluation.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.39 no.3
• /
• pp.141-148
• /
• 2021

This paper presents a method to detect zebra-crossing using deep learning which combines SegNet and ResNet. For the blind, a safe crossing system is important to know exactly where the zebra-crossings are. Zebra-crossing detection by deep learning can be a good solution to this problem and robotic vision-based assistive technologies sprung up over the past few years, which focused on specific scene objects using monocular detectors. These traditional methods have achieved significant results with relatively long processing times, and enhanced the zebra-crossing perception to a large extent. However, running all detectors jointly incurs a long latency and becomes computationally prohibitive on wearable embedded systems. In this paper, we propose a model for fast and stable segmentation of zebra-crossing from captured images. The model is improved based on a combination of SegNet and ResNet and consists of three steps. First, the input image is subsampled to extract image features and the convolutional neural network of ResNet is modified to make it the new encoder. Second, through the SegNet original up-sampling network, the abstract features are restored to the original image size. Finally, the method classifies all pixels and calculates the accuracy of each pixel. The experimental results prove the efficiency of the modified semantic segmentation algorithm with a relatively high computing speed.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.19 no.6
• /
• pp.96-106
• /
• 2020

To solve the limitation of traffic monitoring that occur from embedded sensor such as loop and piezo sensors, the thermal imaging camera was installed on the roadside. As the length of Class 1(passenger car) is getting longer, it is becoming difficult to classify from Class 3(2-axle truck) by using an embedded sensor. The collected images were labeled to generate training data. A total of 17,536 vehicle images (640x480 pixels) training data were produced. CNN (Convolutional Neural Network) was used to achieve vehicle classification based on thermal image. Based on the limited data volume and quality, a classification accuracy of 97.7% was achieved. It shows the possibility of traffic monitoring system based on AI. If more learning data is collected in the future, 12-class classification will be possible. Also, AI-based traffic monitoring will be able to classify not only 12-class, but also new various class such as eco-friendly vehicles, vehicle in violation, motorcycles, etc. Which can be used as statistical data for national policy, research, and industry.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_3
• /
• pp.1837-1846
• /
• 2022

The annual increase in chemical accidents is causing damage to life and the environment due to the spread and residual of substances. Environmental damage investigation is more difficult to determine the geographical scope and timing than human damage investigation. Considering the reality that there is a lack of professional investigation personnel, it is urgent to develop an efficient quantitative evaluation method. In order to improve this situation, this paper conducted a chemical accidents investigation using unmanned aerial vehicles(UAV) equipped with various sensors. The damaged area was calculated by Ortho-image and strength of agreement was calculated using the normalized difference vegetation index image. As a result, the Cohen's Kappa coefficient was 0.649 (threshold 0.7). However, there is a limitation in that analysis has been performed based on the pixel of the normalized difference vegetation index. Therefore, there is a need for a chemical accident investigation plan that overcomes the limitations.

• Journal of the Korean Geotechnical Society
• /
• v.23 no.7
• /
• pp.5-16
• /
• 2007

The monitoring and investigation of underwater landslide help to understand its mechanism, increase the usefuless of design and construction and reduce the losses. This paper presents three high resolution geophysical techniques electrical resisitance, ultrasonic wave reflection imaging, and shear wave tomography conducted to determine the lab-scaled submerged landslide. Electrical resistance profiles of a soil mass obtained by an electrical resistance probe provide detailed information to assess the spatial distribution of the soil mass with milimetric resolution. An ultrasonic wave image obtained by recording the reflections from interfaces of different impedance materials permits detecting layers and landslide with submilimetric resolution. The pixel based image of immersed landslides is created by the inversion of the boundary information achieved from the traveling time of shear waves. The experimental results show that the ultrasonic wave imaging and the electrical resistance can provide complementary information; and their association with S-wave tomography image can produce a 3-D view of the underwater landslide. This study suggests that geophysical techniques may be effective tools for the detection of the underwater landslides and spatial distribution offshore.

• Design Convergence Study
• /
• v.15 no.5
• /
• pp.71-83
• /
• 2016

Interior panels are usually used in finishing of interior walls for not only decorative effects but also information transfer. According to designer's design placing interior panels may need repetitive tasks and the emphasis of this paper is to support an automation of these tasks. Considering the utilization characteristics of interior panels, we propose three method to present patterns by using bitmap image pixels and interior panels' shape changes, based on the theoretical consideration. In addition, in order to approve the possibility of the proposed methods, we have implemented the BIM based interior panels auto layout tool which applied one of the three methods to present patterns by using bitmap image pixel values and panel identification attributes. This tool also supports auto generation of quantity and panel arrangement sequence information that will be used in future construction phase. We expect that this approach will also be used in other decorative objects which require repetition of the basic units, such as floor tiles.

• Journal of Internet Computing and Services
• /
• v.16 no.2
• /
• pp.49-55
• /
• 2015

According to the statistics of traffic accidents over recent 5 years, traffic accidents during the night times happened more than the day times. There are various causes to occur traffic accidents and the one of the major causes is inappropriate or missing street lights that make driver's sight confused and causes the traffic accidents. In this paper, with smartphones, we designed and implemented a lane luminance measurement application which stores the information of driver's location, driving, and lane luminance into database in real time to figure out the inappropriate street light facilities and the area that does not have any street lights. This application is implemented under Native C/C++ environment using android NDK and it improves the operation speed than code written in Java or other languages. To measure the luminance of road, the input image with RGB color space is converted to image with YCbCr color space and Y value returns the luminance of road. The application detects the road lane and calculates the road lane luminance into the database sever. Also this application receives the road video image using smart phone's camera and improves the computational cost by allocating the ROI(Region of interest) of input images. The ROI of image is converted to Grayscale image and then applied the canny edge detector to extract the outline of lanes. After that, we applied hough line transform method to achieve the candidated lane group. The both sides of lane is selected by lane detection algorithm that utilizes the gradient of candidated lanes. When the both lanes of road are detected, we set up a triangle area with a height 20 pixels down from intersection of lanes and the luminance of road is estimated from this triangle area. Y value is calculated from the extracted each R, G, B value of pixels in the triangle. The average Y value of pixels is ranged between from 0 to 100 value to inform a luminance of road and each pixel values are represented with color between black and green. We store car location using smartphone's GPS sensor into the database server after analyzing the road lane video image with luminance of road about 60 meters ahead by wireless communication every 10 minutes. We expect that those collected road luminance information can warn drivers about safe driving or effectively improve the renovation plans of road luminance management.

• Journal of Korean Ophthalmic Optics Society
• /
• v.16 no.4
• /
• pp.409-415
• /
• 2011

Purpose: To design applied anamorphic lens that focal length ratio is 3:1 optical system to improve detecting distance. Methods: We defined a boundary condition as $50^{\circ}{\sim}60^{\circ}$ for viewing angle, horizontal direction 36mm, vertical direction 12 mm for focal length, f-number 4, $15{\mu}m{\times}15{\mu}m$ for pixel size and limit resolution 25% in 33l p/mm. Si, ZnS and ZnSe as a materials were used and 4.8 ${\mu}m$, 4.2 ${\mu}m$, 3.7 ${\mu}m$ as a wavelength were set. optical performance with detection distance, narcissus and athermalization in designed camera were analyzed. Results: F-number 4, y direction 12 mm and x direction 36 mm for focal length of the thermal optical system were satisfied. Total length of the system was 76 mm so that an overall volume of the system was reduced. Astigmatism and spherical aberration was within ${\pm}$0.10 which was less than 2 pixel size. Distortion was within 10% so there was no matter to use as a thermal optical camera. MTF performance for the system was over 25% from 33l p/mm to full field so it was satisfied with the boundary condition. Designed optical system was able to detect up to 2.9 km and reduce a diffused image by decreasing a narcissus value from all surfaces except the 4th surface. From sensitivity analysis, MTF resolution was increased on changing temperature with the 5th lens which was assumed as compensation. Conclusions: Designed optical system which used anamorphic lens was satisfied with boundary condition. an increasing resolution with temperature, longer detecting distance and decreasing of narcissus were verified.

• The Korean Journal of Nuclear Medicine
• /
• v.34 no.4
• /
• pp.322-335
• /
• 2000

Purpose: For quantitative estimation of cerebrovascular reserve (CVR), we estimated the cerebral blood flow (CBF) using Lassen's nonlinearity correction algorithm and Tc-99m HMPAO brain SPECT images acquired with consecutive acquisition protocol. Using the values of CBF in basal and acetaBolamide (ACZ) stress states, CBF increase was calculated. Materials and Methods: In 9 normal subjects (age; $72{\pm}4$ years), brain SPECT was performed at basal and ACZ stress states consecutively after injection of 555 MBq and 1,110 MBq of Tc-99m HMPAO, respectively. Cerebellum was automatically extracted as reference region on basal SPECT image using threshold method. Assuming basal CBF of cerebellum as 55 ml/100g/min, CBF was calculated lot every pixel at basal states using Lassen's algorithm. Cerebellar blood flow at stress was estimated comparing counts of cerebellum at rest and ACZ stress and Lassen's algorithm. CBF of every pixel at ACZ stress state was calculated using Lassen's algorithm and ACZ cerebellar count. CVR was calculated by subtracting basal CBF from ACZ stress CBF for every pixel. The percent CVR was calculated by dividing CVR by basal CBF. The CBF and percentage CVR parametric images were generated. Results: The CBF and percentage CVR parametric images were obtained successfully in all the subjects. Global mean CBF were $49.6{\pm}5.5ml/100g/min\;and\;64.4{\pm}10.2ml/100g/min$ at basal and ACZ stress states, respectively. The increase of CBF at ACZ stress state was $14.7{\pm}9.6ml/100g/min$. The global mean percent CVR was 30.7% and was higher than the 13.8% calculated using count images. Conclusion: The blood flow at basal and ACZ stress states and cerebrovascular reserve were estimated using basal/ACZ Tc-99m-HMPAO SPECT images and Lassen's algorithm. Using these values, parametric images for blood flow and cerebrovascular reserve were generated.

• Korean Journal of Radiology
• /
• v.24 no.8
• /
• pp.807-820
• /
• 2023

Objective: To assess whether computed tomography (CT) conversion across different scan parameters and manufacturers using a routable generative adversarial network (RouteGAN) can improve the accuracy and variability in quantifying interstitial lung disease (ILD) using a deep learning-based automated software. Materials and Methods: This study included patients with ILD who underwent thin-section CT. Unmatched CT images obtained using scanners from four manufacturers (vendors A-D), standard- or low-radiation doses, and sharp or medium kernels were classified into groups 1-7 according to acquisition conditions. CT images in groups 2-7 were converted into the target CT style (Group 1: vendor A, standard dose, and sharp kernel) using a RouteGAN. ILD was quantified on original and converted CT images using a deep learning-based software (Aview, Coreline Soft). The accuracy of quantification was analyzed using the dice similarity coefficient (DSC) and pixel-wise overlap accuracy metrics against manual quantification by a radiologist. Five radiologists evaluated quantification accuracy using a 10-point visual scoring system. Results: Three hundred and fifty CT slices from 150 patients (mean age: 67.6 ± 10.7 years; 56 females) were included. The overlap accuracies for quantifying total abnormalities in groups 2-7 improved after CT conversion (original vs. converted: 0.63 vs. 0.68 for DSC, 0.66 vs. 0.70 for pixel-wise recall, and 0.68 vs. 0.73 for pixel-wise precision; P < 0.002 for all). The DSCs of fibrosis score, honeycombing, and reticulation significantly increased after CT conversion (0.32 vs. 0.64, 0.19 vs. 0.47, and 0.23 vs. 0.54, P < 0.002 for all), whereas those of ground-glass opacity, consolidation, and emphysema did not change significantly or decreased slightly. The radiologists' scores were significantly higher (P < 0.001) and less variable on converted CT. Conclusion: CT conversion using a RouteGAN can improve the accuracy and variability of CT images obtained using different scan parameters and manufacturers in deep learning-based quantification of ILD.