• Title/Summary/Keyword: Gradient operator

Search Result 93, Processing Time 0.035 seconds

Region-Based Gradient and Its Application to Image Segmentation

  • Kim, Hyoung Seok
    • International journal of advanced smart convergence
    • /
    • v.7 no.4
    • /
    • pp.108-113
    • /
    • 2018
  • In this study, we introduce a new image gradient computation based on understanding of image generation. Most images consist of groups of pixels with similar color information because the images are generally obtained by taking a picture of the real world. The general gradient operator for an image compares only the neighboring pixels and cannot obtain information about a wide area, and there is a risk of falling into a local minimum problem. Therefore, it is necessary to attempt to introduce the gradient operator of the interval concept. We present a bow-tie gradient by color values of pixels on bow-tie region of a given pixel. To confirm the superiority of our study, we applied our bow-tie gradient to image segmentation algorithms for various images.

Improved Watershed Image Segmentation Using the Morphological Multi-Scale Gradient

  • Gelegdorj, Jugdergarav;Chu, Hyung-Suk;An, Chong-Koo
    • Journal of the Institute of Convergence Signal Processing
    • /
    • v.12 no.2
    • /
    • pp.91-95
    • /
    • 2011
  • In this paper, we present an improved multi-scale gradient algorithm. The proposed algorithm works the effectively handling of both step and blurred edges. In the proposed algorithm, the image sharpening operator is sharpening the edges and contours of the objects. This operation gives an opportunity to get noise reduced image and step edged image. After that, multi-scale gradient operator works on noise reduced image in order to get a gradient image. The gradient image is segmented by watershed transform. The approach of region merging is used after watershed transform. The region merging is carried out according to the region area and region homogeneity. The region number of the proposed algorithm is 36% shorter than that of the existing algorithm because the proposed algorithm produces a few irrelevant regions. Moreover, the computational time of the proposed algorithm is relatively fast in comparison with the existing one.

An edge detection method for gray scale images based on their fuzzy system representation (디지털 영상의 퍼지시스템 표현을 이용한 Edge 검출방법)

  • 문병수;이현철;김장열
    • Journal of the Korean Institute of Intelligent Systems
    • /
    • v.11 no.6
    • /
    • pp.454-458
    • /
    • 2001
  • Based on a fuzzy system representation of gray scale images, we derive and edge detection algorithm whose convolution kernel is different from the known kernels such as those of Robert's Prewitt's or Sobel's gradient. Our fuzzy system representation is an exact representation of the bicubic spline function which represents the gray scale image approximately. Hence the fuzzy system is a continuous function and it provides a natural way to define the gradient and the Laplacian operator. We show that the gradient at grid points can be evaluated by taking the convolution of the image with a 3$\times$3 kernel. We also that our gradient coupled with the approximate value of the continuous function generates an edge detection method which creates edge images clearer than those by other methods. A few examples of applying our methods are included.

  • PDF

A New Interpretation of the Compass Gradient Edge Operators (Compass Gradient Edge 연산자의 새로운 해석방법)

  • Park, Rae-Hong;Choi, Woo Young
    • Journal of the Korean Institute of Telematics and Electronics
    • /
    • v.24 no.1
    • /
    • pp.97-101
    • /
    • 1987
  • The edge, a discontinuity or abrupt change in the gray-level or color, is a fundamentally important primitive feature of an image necessary for the image analysis and classification. Two-dimensional 3x3 compass gradient operators (ex. Sobel, Prewitt, and Kirsch operators)are commonly used in the edge detection and usually detect 8 compass directional components. In this paper, we present a new interpretation of the relationships between the resulting 8 gradient magnitudes and the 8 intensity values of neighboring pixels which are covered by the two-dimensional 3x3 mask. It is expected that a new gradient edge operator may be designed by changing the eigenvalues in the transform domain and the fast optical edge operator may be implemented by using the optical system.

  • PDF

Edge Detection Using the Information of Edge Structural Regions (에지의 구조적 영역정보를 이용한 에지검출)

  • 김수겸;박중순;최정희
    • Journal of Advanced Marine Engineering and Technology
    • /
    • v.24 no.2
    • /
    • pp.82-89
    • /
    • 2000
  • Edge detection is the first step and very important step in image analysis. In this paper, proposed edge detection operators based on informations of edge types and it is different from other classical edge detection operators such as gradient and surface fitting operators. The first, we defined characteristics of edge types such as localization, thinness, length. The second, we defined valid edge types and ideal edge pixel positions in $3\times3$window based on edge characteristics of edge types. And we proposed edge detection algorithm and twelve windows based on valid edge types. In specially, proposed algorithm was shown better performence of edge detection than other operators such as gradient operator and the LoG(Laplacian of Gaussian) operator of zero crossings.

  • PDF

Context-free multiple-object segmentation using attention operator based on modified generalized symmetry transform (일반화 대칭변환을 변형한 관심 연산자에 의한 사전 정보없는 다중 물체 분할)

  • 구태모;전준형;최흥문
    • Journal of the Korean Institute of Telematics and Electronics C
    • /
    • v.34C no.4
    • /
    • pp.36-44
    • /
    • 1997
  • An efficient context-free multiple-object segmentation using attention operator based on modified generalized symmetry transform is proposed and implemented by modifying a radial basis function network. By using the difference of intensity gradient, instead of te intensity gradient itself, in generalized symmetry tranform so as to make the attention operator to preserve the edges of the objects shape, an efficient context-free multiple-object segementation is proposed in which no a priori shape informtion on the objects is requried. The attention operator is implemented by using a modified radial basis function network which can reflect symmetry, and by using te edge pyramid of the input image, both of the local and the global symmetry of the objects are reflected simultaneously to make the multiple-object with different sizes be segmented with a singel fixed-size $n\timesm$ can be done with O(n) complexity. The simulaton results show that the proposed algorithm can efficiently be used in context-free multiple-object segmentation even for the low contrast IR images as well as for the images from the camera.

  • PDF

SUPERCONVERGENT GRADIENT RECOVERY FOR THE PARABOLIC INITIAL BOUNDARY VALUE PROBLEM

  • LAKHANY, AM;WHITEMAN, JR
    • Journal of the Korean Society for Industrial and Applied Mathematics
    • /
    • v.3 no.1
    • /
    • pp.1-15
    • /
    • 1999
  • Gradient recovery techniques for the second order elliptic boundary value problem are well known. In particular, the Midpoint and the Vertex Recovery Operator have been studied by various authors and under suitable assumptions on the regularity of the unknown solution superconvergence property of these recovered gradients have been proved. In this paper we extend these results to the recovered gradient of the finite element approximation to a model initial-boundary value problem, and go on to prove superconvergence result for this recovered gradient in a discrete (in time) error norm.

  • PDF

Palmprint Verification Using Multi-scale Gradient Orientation Maps

  • Kim, Min-Ki
    • Journal of the Optical Society of Korea
    • /
    • v.15 no.1
    • /
    • pp.15-21
    • /
    • 2011
  • This paper proposes a new approach to palmprint verification based on the gradient, in which a palm image is considered to be a three-dimensional terrain. Principal lines and wrinkles make deep and shallow valleys on a palm landscape. Then the steepest slope direction in each local area is first computed using the Kirsch operator, after which an orientation map is created that represents the dominant slope direction of each pixel. In this study, three orientation maps were made with different scales to represent local and global gradient information. Next, feature matching based on pixel-unit comparison was performed. The experimental results showed that the proposed method is superior to several state-of-the-art methods. In addition, the verification could be greatly improved by fusing orientation maps with different scales.

The Selection of Significant Points from Grid DEM by High-Pass Filtering (하이-패스 필터링에 의한 격자형 수치표고모델의 중심점 추출)

  • 이석찬;최병길
    • Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
    • /
    • v.9 no.2
    • /
    • pp.139-149
    • /
    • 1991
  • In general, Digital Elevation Models(DEM) are constructed in a grid format for this form is advantageous to capture data automatically and is easy to manipulate. But, grid DEM requires vast volumes of data to represent terrain features finely and accurately as its data is sampled in a regular space. This paper aims at constructing compact DEM by selecting the significant points from grid DEM which affect well the terrain features. For the purpose, the significant points is detected by the high-pass filtering using Laplacian operator and gradient operator. The results of this study show that the Laplacian operator is more efficient than the gradient operator in selecting the significant points for compact DEM.

  • PDF

Texture Feature Analysis of Machined Surface Image Using Intensity Gradient (광 강도변화를 이용한 가공면 영상의 텍스쳐 특징분석)

  • 사승윤
    • Journal of the Korean Society of Manufacturing Technology Engineers
    • /
    • v.7 no.6
    • /
    • pp.49-56
    • /
    • 1998
  • Super precision working technique and machine tool have been continually developed thanks to advanced electronic field. To obtain good result. it is necessary to investigate surface in grinding with $mu extrm{m}$ level. There were quite many researches to satisfy these demands by using non-contact methods through the computer vision. In this study, the texture of working surface was analyzed. co-occurrence matrices was obtained from the surface roughness. Texture parameter was obtained using position operator composed of $ heta$, d according to variation of angle direction and distance. As a result, it was found that surface texture was more affected by direction($\theta$) than distance(d).

  • PDF