Unlock instant, AI-driven research and patent intelligence for your innovation.

Method for Extracting Same Name Points from Images Based on Local Orientation Histogram Descriptor

A technology of direction histogram and extraction method, applied in the field of image processing, can solve the problem of limited distinguishing power, and achieve the effect of enhancing and improving distinguishing power

Active Publication Date: 2018-07-03
CHINA CENT FOR RESOURCES SATELLITE DATA & APPL
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Therefore, for SAR image matching, the discriminative power of the above descriptors is limited

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
  • Method for Extracting Same Name Points from Images Based on Local Orientation Histogram Descriptor
  • Method for Extracting Same Name Points from Images Based on Local Orientation Histogram Descriptor
  • Method for Extracting Same Name Points from Images Based on Local Orientation Histogram Descriptor

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0028] For SAR image matching, the characteristics of high similarity between scene features should be fully considered, and emphasis should be placed on enhancing the discriminative power of descriptors.

[0029] SIFT divides the feature point neighborhood into 4*4 sub-blocks, counts the gradient direction histogram of each sub-block and concatenates them to form a description sub-vector. The gradient direction has a certain degree of discrimination, but the gradient direction is the direction in which the pixel gray value changes the most. If only the maximum direction is used and other directions are ignored, the discrimination will be insufficient. In view of this, the present invention proposes a new descriptor based on Local Sorted Orientation Histogram (LSOH), which first calculates the directional derivatives of each point in eight directions, and then takes the largest and second largest The direction response of the point is used as a description of the point, and fi...

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 image homonym point extraction method based on the local sorting direction histogram descriptor, aiming at the problem that the SIFT block gradient direction histogram descriptor has insufficient discrimination due to the high similarity of SAR image scene features, proposes a new method based on the local sorting direction Descriptor for the histogram. The method of the invention calculates the directional derivatives of eight directions of each pixel point in the characteristic neighborhood, takes the direction positions of the largest and second largest derivatives as the characteristic primitives, divides the neighborhood into blocks and counts the distribution of the characteristic primitives to form a descriptor vector. By adding the sub-major direction and its location information, more complete neighborhood information is considered, and the discriminative power of the descriptor is improved. Experimental results in scenes such as mountainous areas, cities, and villages show that the method of the present invention can better measure the differences between features, enhance the discrimination of descriptors, and can extract more matching points, which is more suitable for SAR image matching.

Description

technical field [0001] The invention belongs to the field of image processing, and relates to a method for extracting homonymous points of remote sensing images, which is suitable for SAR image registration, splicing, block adjustment and the like. Background technique [0002] SAR image homonym point extraction is to extract homonym feature points from two SAR images with geometric and grayscale distortions, which is a key technology for many applications such as image registration, image mosaic, target detection and recognition. [0003] In recent years, the method based on invariant features has been a research hotspot in the automatic extraction of points with the same name. This type of method is divided into four stages: feature detection, feature description, feature matching, and feature screening: first, the detector is used to extract stable points and lines in the image. Or surface features, and then extract the structure, shape, texture and other information of t...

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): G06T7/33G06K9/46
CPCG06T2207/10044G06V10/462
Inventor 王山虎郝雪涛王峰吕江安
Owner CHINA CENT FOR RESOURCES SATELLITE DATA & APPL