• The Journal of Korean Institute of Information Technology
• /
• v.16 no.12
• /
• pp.75-83
• /
• 2018

Recently, autonomous vehicles have been actively studied. Pedestrian detection and recognition technology is important in autonomous vehicles. Pedestrian detection using CNN(Convolutional Neural Netwrok), which is mainly used recently, generally shows good performance, but there is a performance degradation depending on the environment of the image. In this paper, we propose a pedestrian detection system applying long-term memory structure of hippocampal neural network based on CNN network with LGP-FL (Local Gradient Pattern-Feature Layer) added. First, change the input image to a size of $227{\times}227$. Then, the feature is extracted through a total of 5 layers of convolution layer. In the process, LGP-FL adds the LGP feature pattern and stores the high-frequency pattern in the long-term memory. In the detection process, it is possible to detect the pedestrian more accurately by detecting using the LGP feature pattern information robust to brightness and color change. A comparison of the existing methods and the proposed method confirmed the increase of detection rate of about 1~4%.

• Progress in Medical Physics
• /
• v.29 no.4
• /
• pp.150-156
• /
• 2018

This study aims to develop an improved Feldkamp-Davis-Kress (FDK) reconstruction algorithm using anisotropic total variation (ATV) minimization to enhance the image quality of low-dose cone-beam computed tomography (CBCT). The algorithm first applies a filter that integrates the Shepp-Logan filter into a cosine window function on all projections for impulse noise removal. A total variation objective function with anisotropic penalty is then minimized to enhance the difference between the real structure and noise using the steepest gradient descent optimization with adaptive step sizes. The preserving parameter to adjust the separation between the noise-free and noisy areas is determined by calculating the cumulative distribution function of the gradient magnitude of the filtered image obtained by the application of the filtering operation on each projection. With these minimized ATV projections, voxel-driven backprojection is finally performed to generate the reconstructed images. The performance of the proposed algorithm was evaluated with the catphan503 phantom dataset acquired with the use of a low-dose protocol. Qualitative and quantitative analyses showed that the proposed ATV minimization provides enhanced CBCT reconstruction images compared with those generated by the conventional FDK algorithm, with a higher contrast-to-noise ratio (CNR), lower root-mean-square-error, and higher correlation. The proposed algorithm not only leads to a potential imaging dose reduction in repeated CBCT scans via lower mA levels, but also elicits high CNR values by removing noisy corrupted areas and by avoiding the heavy penalization of striking features.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.13 no.6
• /
• pp.568-574
• /
• 2020

In this paper, we propose a method to reduce the square root computation having high computation complexity in Canny edge detection algorithm using image processing. The proposed method is to reduce the number of operation calculating gradient magnitude using pixel's continuity using make a specific pattern instead of square root computation in gradient magnitude calculating operation. Using various test images and changing number of hole pixels, we can check for calculate match rate about 97% for one hole, and 94%, 90%, 88% when the number of hole is increased and measure decreasing computation time about 0.2ms for one hole, and 0.398ms, 0.6ms, 0.8ms when the number of hole is increased. Through this method, we expect to implement low power embedded vision system through high accuracy and a reduced operation number using two-hole pixels.

• Economic and Environmental Geology
• /
• v.30 no.6
• /
• pp.625-635
• /
• 1997

A back projection algorithm is applied to 216 Pn travel time measurements to image lateral variations of compressional velocity in the uppermost mantle in the Korean Peninsula. We obtained an average P-velocity value for the uppermost mantle of $7.90{\pm}0.18km/sec$, and an average mantle P-velocity gradient of $5.3{\times}10^{-3}s^{-1}$ for the Korean Peninsula. The final 3-D velocity image in the uppermost mantle is characterized by a low-velocity (about $7.77{\pm}0.12km/sec$) region in the southeast area of the Korean peninsula, which is called 'Kyongsang Basin' and by high-velocity(${\geq}8.08km/sec$) region in the northern area of the Korean Peninsula(Hamkyong and Pyongan provinces). The crustal thicknesses are calculated for the 10 subregions. The crustal thickness of the northern part(${\geq}39^{\circ}N$) of the Korean Peninsula is 33.0-36.0 km, on the contrary, that of the southern part(< $39^{\circ}N$) is 30.7~33.7 km. The velocity image obtained in this study is somewhat consistent with previous S-P travel time studies and gravity studies.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.19 no.5
• /
• pp.672-677
• /
• 2010

The goal of this study is to simplify the measurement process of the flat zone length produced by a flexible disk grinding system for the process automation. The image of workpiece in the grinding process is obtained, and the cutting speed and the feeding speed are controlled carefully to maximize the flat zone length. The gradient, the inflection point and the length of the line in the image are calculated, and the length is also measured by using a projector. Processing conditions and inversely proportional to flat zone length was changing. The flat zone length is estimated by Neural network algorithm considering the process conditions with the estimated error range as 0.06~3.61%, the Neural network algorithm for the grinding process estimation is found to be useful for building the process automation database.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.5
• /
• pp.413-418
• /
• 2001

In this paper, we propose a new camera calibration method which is based on a fuzzy model instead of a physical camera model of the conventional method. The camera calibration is to determine the correlation between camera image coordinate and real world coordinate. The camera calibration method using a fuzzy model can not estimate camera physical parameters which can be obtained in the conventional methods. However, the proposed method is very simple and efficient because it can determine the correlation between camera image coordinate and real world coordinate without any restriction, which is the objective of camera calibration. With calibration points acquired out of experiments, 3-D real world coordinate and 2-D image coordinate are estimated using the fuzzy modeling method and the results of the experiments demonstrate the validity of the proposed method.

• Journal of Institute of Control, Robotics and Systems
• /
• v.3 no.5
• /
• pp.497-502
• /
• 1997

In this study an automatic rolling-coil labeling system using robot vision system and peripheral mechanism is proposed and implemented, which instead of the manual labor to attach labels Rolling-coils in a steel mill. The binary image process for the image processing is performed with the threshold, and the contour line is converted to the binary gradient which detects the discontinuous variation of brightness of rolling-coils. The moments invariant algorithm proposed by Hu is used to make it easy to recognize even when the position of the center are different from the trained data. The position error compensation algorithm of six degrees of freedom industrial robot manipulator is also developed and the data of the position of the center rolling-coils, which is obtained by floor mount camera, are transferred by asynchronous communication method. Therefore, even if the position of center is changed, robot moves to the position of center and performs the labeling work successfully. Therefore, this system can be improved the safety and efficiency.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.9
• /
• pp.930-935
• /
• 2014

A Gabor filter is a linear filter used for edge detectionas frequency and orientation representations of Gabor filters are similar to those of the human visual system. In this thesis, we propose a pedestrian detection algorithm using a Gabor filter bank. In order to extract the features of the pedestrian, we use various image processing algorithms and data structure algorithms. First, color image segmentation is performed to consider the information of the RGB color space. Second, histogram equalization is performed to enhance the brightness of the input images. Third, convolution is performed between a Gabor filter bank and the enhanced images. Fourth, statistical values are calculated by using the integral image (summed area table) method. The calculated statistical values are used for the feature matrix of the pedestrian area. To evaluate the proposed algorithm, the INRIA pedestrian database and SVM (Support Vector Machine) are used, and we compare the proposed algorithm and the HOG (Histogram of Oriented Gradient) pedestrian detector, presentlyreferred to as the methodology of pedestrian detection algorithm. The experimental results show that the proposed algorithm is more accurate compared to the HOG pedestrian detector.

• The Journal of the Korea Contents Association
• /
• v.3 no.4
• /
• pp.17-22
• /
• 2003

In this paper, we propose the new side-view face detection method in color images which contain faces over one. It uses color and the geometrical distance between nose and chin. We convert RGB to YCbCr color space. We extract candidate regions of face using skin color information from image. And then, the extracted regions are processed by morphological filter, and the processed regions are labeled. Also, we correct the gradient of inclined face image using projected character of nose. And we detect the inclined side-view faces that have right and left 45 tips by within via ordinate. And we get 92% detection rate in 100 test images.

• ETRI Journal
• /
• v.38 no.5
• /
• pp.996-1007
• /
• 2016

This paper presents new hole-filling methods for generating multiview images by using depth image based rendering (DIBR). Holes appear in a depth image captured from 3D sensors and in the multiview images rendered by DIBR. The holes are often found around the background regions of the images because the background is prone to occlusions by the foreground objects. Background-oriented priority and gradient-oriented priority are also introduced to find the order of hole-filling after the DIBR process. In addition, to obtain a sample to fill the hole region, we propose the fusing of depth and color information to obtain a weighted sum of two patches for the depth (or rendered depth) images and a new distance measure to find the best-matched patch for the rendered color images. The conventional method produces jagged edges and a blurry phenomenon in the final results, whereas the proposed method can minimize them, which is quite important for high fidelity in stereo imaging. The experimental results show that, by reducing these errors, the proposed methods can significantly improve the hole-filling quality in the multiview images generated.