• Journal of KIISE:Software and Applications
• /
• v.34 no.11
• /
• pp.961-970
• /
• 2007

Since the usual conventional edge detectors employ the local differential derivatives, the detected edges are not uniform in their widths or some edges are missed out of the detection on magnified images. We employ a mapping from the exactly monotonic intensity distributions of ramp edges to the simple step functions of intensity, which is referred to as perfect sharpening map of ramp edges. This map is based on the non-local feature of intensity distribution and used to introduce a modified differentiation, in terms of which we can construct an efficient edge detector adaptive to the variation of edge width. By adopting the operator MADD in this paper, we developed an edge detector that works stably against the magnification of image or the variation of edge width. It is shown by comparing to the conventional algorithms that the proposed one is very excellent.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.9 no.3
• /
• pp.143-154
• /
• 2001

An algorithm for estimating the lane departure of a vehicle is derived and implemented based on an EDF(edge distribution function) obtained from gray-level images taken by a CCD camera mounted on a vehicle. As the function of edge direction, the EDF is aimed to show the distribution of edge direction and to estimate the possibility of lane departure with respect to its symmetric axis and local mamma. The EDF plays important roles: 1) It reduces noisy effects caused by dynamic road scene. 2) It makes possible lane identification without camera modeling. 3) It also leads LDW(lane departure warning) problem to a mathematical approach. When the situations of lane departure such that the vehicle approaches to lane marks or runs in the vicinity of the lane marks are occurred, the orientation of lane marks in images is changed, and then the situations are immediately reflected to the EDF. Accordingly, the lane departure is estimated by studying the shape of the EDF. The proposed EDF-based algorithm enhanced the adaptability to cope with the random and dynamic road environments, and eventually led to the reliable LDW system.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2015.10a
• /
• pp.921-923
• /
• 2015

Edge in the image includes the size, direction and location of objects. The existing detection methods for detecting this edge is a method using Sobel, Prewitt, Roberts and Laplacian, etc. These existing methods use a fixed weighted mask in order to detect the edge and have somewhat insufficient edge detection characteristics. Therefore in this paper, an algorithm that detects the edge by applying the grey level converting function according to the pixel distribution of local mask was proposed.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2015.07a
• /
• pp.45-48
• /
• 2015

오늘날 하드웨어의 발전으로 인해 영상 해상도는 FHD를 넘어 4K UHD 이상의 영상 해상도가 사용화되고 있다. 하지만 Edge Map을 만들기 위해 일반적으로 사용하는 함수들은 Convolution 함수 일종으로서 영상의 해상도가 높을수록 더 많은 Complexity를 요구한다. 또한 현재 주요 영상 압축 기술인 JPEG, H.264/AVC High efficiency video coding(HEVC)같은 기법들은 자연 영상을 중점으로 개발되어 있어 Edge map 압축에 있어 자연 영상만큼의 효율을 보여주지 못하고 있다. 본 논문은 원 영상을 Down Scaling한 뒤 이미지를 다시 원래 사이즈로 Up Scaling하여 두 영상의 차를 이용한 Edge Map을 생성하는 새로운 방법을 소개한다. 생성된 Edge Map의 특성인 Histogram 값의 분포가 0을 중심으로 Gaussian 분포를 가지는 것을 이용한 Zero Based 코덱을 제안한다. 제안된 알고리즘을 이용하여 고 해상도 영상에서도 빠르게 Edge Map을 생성하고 제안한 코덱을 통해 해당 Edge map을 압축한 결과 다른 압축 기술보다 더 뛰어난 성능을 보여주었다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.10 no.3
• /
• pp.549-555
• /
• 2006

An edge detection is an very important area in image processing and computer vision, General edge detection methods (Robert mask, Sobel mask, Kirsh mask etc) are a good performance to detect step edge in a image but are no good performance to detect step edge in a noses image. We suggested a step edge detection method based on image information measure and neutral network. Using these essential properties of step edges, which are directional and structural and whose gray level distribution in neighborhood, as a input vector to the BP neutral network we get the good result of proposed algorithm. And also we get the satisfactory experimental result using rose image and cell images an experimental and analysing image.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.14 no.2
• /
• pp.513-520
• /
• 2010

The degradation phenomenon caused by noises significantly corrupts digitalized data. Therefore, a variety of methods to preserve the edge component of signals and remove noise simultaneously have been used in time domain and frequency domain. In this paper, we have proposed a new noise reduction algorithm using wavelet approximation coefficients to reduce the mixed noise overlapping the signal. The proposed algorithm adopts the distribution characteristics of the error function which is obtained by accumulating the wavelet approximation coefficients, in order to improve the capability to separate edges of the signal and noises.

• The Korean Journal of Applied Statistics
• /
• v.32 no.5
• /
• pp.721-740
• /
• 2019

Noise reduction is an important field in image processing and requires a statistical approach. However, it is difficult to assume a specific distribution of noise, and a spatial filter that reflects regional characteristics is a small sample and cannot be accessed in a parametric manner. The first order image differential and the second order image differential show a clear difference according to the noise level included in the image and can be more clearly understood using the canyon edge detector. The Fligner-Killeen test was performed and the bootstrap method was used to statistically check the noise level. The estimated noise level was set between 0 and 1 using the cumulative distribution function of the beta distribution. In this paper, we propose a nonparametric noise reduction algorithm that accounts for the noise level included in the image.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.10 no.5
• /
• pp.147-159
• /
• 2002

This paper proposes a method to recognize the noisiness of road images connected with the extraction of lane-related information in order to prevent the usage of erroneous information. The proposed method uses a fuzzy neural network(FNN) with the back-Propagation loaming algorithm. The U decides road images good or bad with respect to visibility of lane marks on road images. Most input parameters to the FNN are extracted from an edge distribution function(EDF), a function of edge histogram constructed by edge phase and norm. The shape of the EDF is deeply correlated to the visibility of lane marks of road image. Experimental results obtained by simulations with real images taken by various lighting and weather conditions show that the proposed method was quite successful, providing decision-making of noisiness with about 99%.

• Convergence Security Journal
• /
• v.8 no.4
• /
• pp.127-134
• /
• 2008

In this paper, we try to improve the accuracy of automatic measurement for analysis equipment by detecting efficiently the edge of a waterdrop with transparency. In order to detect the edge of a waterdrop with transparency, we use an edge detecting technique, MADD (Modified Adaptive Directional Derivative), which can identify the ramp edges with various widths as the perfectly sharp edges and respond effectively regardless of enlarging or reducing the image. The proposed edge detecting technique by means of perfect sharpening of ramp edges employs the modified adaptive directional derivatives instead of the usual local differential operators in order to detect the edges of image. The modified adaptive directional derivatives are defined by introducing the perfect sharpening map into the adaptive directional derivatives. Finally we apply the proposed method to contact angle arithmetic and show the effiency and validity of the proposed method.

• The Korean Journal of Applied Statistics
• /
• v.24 no.2
• /
• pp.335-345
• /
• 2011

In this paper, we propose a new hybrid speed function for image segmentation using level set. A new proposed speed function uses the region and boundary information of image object for the exact result of segmentation. The region information is defined by the probability information of pixel intensity in a ROI(region-of-interest), and the boundary information is defined by the gradient vector flow obtained from the gradient of image. We show the results of experiment for an various artificial image and real medical image to verify the accuracy of segmentation using proposed method.