Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

An Anomaly Detection Method Based on Edge Detection and Color Matching

An anomaly detection and edge detection technology, applied in the field of computer vision, which can solve the problems of insufficient robustness, insufficient stability, unsuitable for large-scale, long-time interval change detection, etc.

Active Publication Date: 2020-10-30
SUN YAT SEN UNIV +1
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to solve the technical defects of insufficient robustness and insufficient stability of the anomaly detection method based on image element comparison provided by the prior art, and to solve the shortcomings of the anomaly detection method based on foreground and background segmentation provided by the prior art. Technical flaws suitable for unattended large-scale, long-time interval change detection, providing an anomaly detection method based on edge detection and color matching

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • An Anomaly Detection Method Based on Edge Detection and Color Matching
  • An Anomaly Detection Method Based on Edge Detection and Color Matching
  • An Anomaly Detection Method Based on Edge Detection and Color Matching

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0024] Such as figure 2 As shown, the anomaly detection method based on edge detection and color matching provided by the present invention includes three steps of image registration, color balance, and anomaly detection.

[0025] 1. Image registration

[0026] The two images taken by the UAV at different times are different in scale, angle, range, etc. The purpose of image registration is to align the corresponding positions of the two images by performing a specific affine transformation.

[0027] Such as image 3 As shown, the image registration is divided into three stages: (1) detection and matching of SIFT feature points; (2) screening of matching pairs of feature points; (3) selection of key points and image registration.

[0028] (1) Detection and matching of SIFT feature points

[0029] 1) Detect SIFT feature points from two images A and B, and obtain corresponding feature vectors;

[0030] 2) For each SIFT feature point in image A, use brute force matching to pe...

Embodiment 2

[0071] In order to verify the beneficial effect of the method provided by the present invention, a specific experiment was carried out in this embodiment. The data set used in this experiment is the SZTAKI data set, which contains 12 pairs of color images, corresponding to parts of the Hungarian cities Szada and Tiszadob. The size of each pair of images is 952×640 pixels with a resolution of 1.5 m / pixel, and the images were taken at five and seven-year intervals, respectively. The changing areas in the image include new urban areas, building construction, tree planting, cultivated land reclamation, etc.

[0072] In the experiment, the method of only edge anomaly detection and the method of edge anomaly detection and color anomaly detection were tested respectively. The results are shown in Table 1:

[0073] Table 1: Comparison of experimental results between the traditional method using only edge detection and the method proposed by the present invention

[0074]

[0075] ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The present invention proposes an abnormality detection method based on edge detection and color matching, through registration, color balance, edge detection and color matching of two aerial images collected at different times, to efficiently and accurately detect abnormal areas therein. Aiming at the deficiencies of the traditional anomaly detection method based on image element comparison, the present invention introduces neighborhood-based feature point matching pair screening and a key point selection strategy based on distance and area criteria in the image registration step, which improves the reliability of image registration. The combination of edge detection and color matching in the image comparison process solves the problem of conflict between anomaly detection and color balance, and improves the accuracy of anomaly detection.

Description

technical field [0001] The present invention relates to the technical field of computer vision, and more specifically, to an anomaly detection method based on edge detection and color matching. Background technique [0002] In the field of surface environment monitoring and other fields, it is necessary to collect images to compare the changes of the same area over time, and to find and mark abnormal areas. For example, use drones to take images of the same place at different times, and conduct comparative analysis to detect abnormal changes in the images, such as illegal buildings, vegetation destruction, and river pollution. At present, there are mainly two types of anomaly detection methods, one is based on image element comparison, and the other is based on foreground and background segmentation. [0003] The anomaly detection method based on the comparison of image elements extracts image elements such as points, lines, and blocks from the image, and then compares them...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/00G06K9/46G06K9/62
CPCG06V20/13G06V10/462G06V10/751
Inventor 郑慧诚贾笑卿李烨
Owner SUN YAT SEN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products