China
Africa
Explore chapters and articles related to this topic
Reverse Engineering and Inspection of Machined Parts in Manufacturing Systems
Published in Cornelius Leondes, Computer-Aided Design, Engineering, and Manufacturing, 2019
Tarek Sobh, Jonathan Owen, Mohamed Dekhil
Slope is taken to be the orientation of the gradient of the difference of Gaussians (DOG) function. The DOG function is an approximation to the Laplacian as mentioned in the Zero-Crossings section of this chapter. The derivatives of slope are computed using a forward difference technique, and the results are smoothed a user-controlled number of times. A graph of curvature vs. distance along a curve can be seen in Fig. 6.36. For each arc segment, a circle is fitusing a least-squares fit [14], and then the endpoints of the arc segment are grown until the distance from the contour to the fitted circle exceeds a tolerance. This process is repeated until growing has no effect or another segment is reached. A similar method is used for the line segments. Segment data is stored as a linked list (see Fig. 6.37).
Introduction to Visual Computing
Published in Ragav Venkatesan, Baoxin Li, Convolutional Neural Networks in Visual Computing, 2017
SIFT is an algorithm for detecting feature points (customarily called SIFT features) that are supposedly invariant to changes in scale and rotation, or have a slight appearance change due to varying illumination or local geometric distortion. The algorithm first creates a multiscale representation of the original image. Then it finds the extrema in the difference of Gaussians in that representation and uses them as potential feature points. Some criteria are used to discard any potential feature point that is deemed a poor candidate (e.g., low-contrast points). Finally, a descriptor will be computed for any of the remaining points and its neighboring region. This descriptor (and its location in an image) is basically what we call a SIFT feature. The full algorithm as described in Lowe (1999) also covers other implementation issues related to fast matching of objects based on SIFT features.
A Comparative Study of Illumination Invariant Techniques in Video Tracking Perspective
Published in IETE Technical Review, 2020
C. S. Asha, A. V. Narasimhadhan
Difference of Gaussian is a band pass filter applied to an input image in the logarithmic domain [27]. Let represents low pass Gaussian filter with standard deviation . Let represents another low pass Gaussian filter with standard deviation . The DoG-filtered output is obtained by convolving an input image with . The parameters and are used for DoG filter. Application of DoG filter on shaking video frames is shown in Figure 8. Since DoG is a bandpass filter, it removes the low-frequency illumination and retains the edge details.