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

Improved feature extraction method and image splicing method based on same

A technology of feature extraction and image stitching, applied in image enhancement, image analysis, image data processing, etc., can solve the problems of relatively high image overlap and size requirements, unsuitable for UAV image stitching, and high algorithm complexity. Ensure the accuracy of image stitching, reduce the time required for stitching, and achieve low computational complexity

Pending Publication Date: 2021-11-30
HARBIN INST OF TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In 2011, Rublee et al. improved the FAST feature detector and the BRIEF feature descriptor, and obtained an operator ORB (Oriented FAST and Rotated BRIEF) that can be calculated quickly and perform well under different lighting conditions and rotation changes. )(Bay H,TuytelaarsT,Van Gool L.SURF:Speeded up Robust Features[C].European conference on computer vision,2006:404-417.), but the operator performs poorly in the scene where the scale changes
[0005] The feature-based image registration algorithm can guarantee the quality of the image and has good robustness, but the algorithm has high complexity and a large amount of calculation; the region-based registration algorithm can well solve the problem of image rotation, The problem of translation and zooming, however, this type of algorithm has relatively high requirements on image overlap and size, and is not suitable for splicing UAV images

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
  • Improved feature extraction method and image splicing method based on same
  • Improved feature extraction method and image splicing method based on same
  • Improved feature extraction method and image splicing method based on same

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0054] Example 1: An improved region-based feature extraction method

[0055] There are many ways to achieve image stitching, and the specific details of different algorithms are different to some extent, but the implementation steps involved are roughly the same. In general, image stitching mainly follows the figure 1 The process shown proceeds.

[0056] The general steps of image stitching are as follows:

[0057] (1) Feature extraction

[0058] By analyzing the image and finding the solution that satisfies the corresponding extremum conditions, the position coordinates of the feature points in the image are obtained. At the same time, in order to describe the feature point, the corresponding feature descriptor is constructed as the description vector of the feature point. For the needs of subsequent related work, the extracted features must remain unchanged under the conditions of uneven illumination, image translation, rotation and scaling. At present, the feature dete...

Embodiment 2

[0108] Embodiment 2: An image mosaic method based on the feature extraction method described in Embodiment 1

[0109] (1) Image preprocessing

[0110] Input the image to be spliced ​​I A , I B , respectively for I A , I B Perform image rotation, image enhancement, and smoothing preprocessing.

[0111] (2) ORB feature extraction

[0112] Obtain binary feature strings according to the feature extraction method described in Embodiment 1.

[0113] (3) Eliminate mismatches

[0114] Obtain feature point pairs through the k-nearest neighbor algorithm, and then filter I through the random consistent sampling algorithm (RANSAC) A and I B The feature points to be matched are used to eliminate a large number of false matches; at the same time, the Euclidean distance between feature descriptors is used as the main reference for feature registration, and the feature points with better matching effect are selected by setting the threshold t. For the image to be stitched For each fe...

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 invention discloses an improved feature extraction method and an image splicing method based on the same, and belongs to the technical field of digital image processing. According to the invention, the splicing speed of aerial images is improved under the condition that the splicing precision is not reduced. According to the feature extraction method, feature extraction is performed on to-be-spliced images I and I by using an FAST-9 algorithm, feature points are obtained by using Harris corner detection, a feature string is obtained through an improved BRISK algorithm, and the feature extraction is realized. Based on an ORB image splicing algorithm, the image splicing method combines characteristics of aerial images returned by an unmanned aerial vehicle to provide an improved ORB-based rapid image splicing algorithm, and realizes the technical effect of rapidly and efficiently obtaining a panorama. Through verification by simulation experiments, the invention can effectively improve the splicing speed of various splicing algorithms under the condition that the precision is not lost; the invention can be widely applied to splicing processing of aerial images; and the invention can quickly and accurately obtain global images of aerial photographing areas.

Description

technical field [0001] The invention belongs to the technical field of digital image processing, and in particular relates to an improved feature extraction algorithm and an image splicing algorithm based on the algorithm. Background technique [0002] In the harsh environment after the disaster, it becomes dangerous and slow to collect image information on the spot, and the images acquired by ordinary cameras often have a relatively small field of view. When the shooting scene becomes larger, the resolution of the images obtained will be very low. Often results in distortion. UAV shooting has the characteristics of flexibility and mobility, which is very suitable for on-site image collection after a disaster occurs. When a disaster occurs, it is very important to quickly obtain the global image information of the disaster-affected area and grasp the distribution of the disaster situation in the area from the global perspective for the follow-up post-disaster rescue work. ...

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 Applications(China)
IPC IPC(8): G06K9/62G06K9/46G06T5/50G06T7/33
CPCG06T5/50G06T7/33G06T2207/20221G06F18/2113
Inventor 石振锋张萌菲张孟琦
Owner HARBIN INST OF TECH
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

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com