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

PCA-SIFT-based fast image splicing method

An image stitching and fast technology, applied in the field of image processing, can solve the problems of high time overhead, not considering the distribution of feature points, difficult to meet real-time performance, etc., and achieve the effect of excellent matching rate

Active Publication Date: 2018-09-14
FUZHOU UNIV
View PDF9 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this algorithm does not consider the distribution of feature points, and too many feature points will be extracted in areas with complex details. This disadvantage is that it is easy to cause mismatches, and the SIFT algorithm is complex to calculate and time-consuming, and it is difficult to meet real-time performance.

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
  • PCA-SIFT-based fast image splicing method
  • PCA-SIFT-based fast image splicing method
  • PCA-SIFT-based fast image splicing method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0042] Embodiment 1, as shown in Figure 2(a), adopts the lena picture as the experimental image of dispersion analysis, as shown in Figure 2(b) and Figure 2(c), the SIFT algorithm can extract 1127 features in the image points, the dispersion of feature points is 126.628, the improved method of the present invention can extract 701 feature points, and the dispersion is 133.706. For the distribution of feature points in an image. The larger the dispersion S, the more discrete and uniform the distribution of feature points is; the smaller S is, the denser and more uneven the distribution of feature points is. It can be found from Fig. 2 that the feature point distribution of the picture processed by the improved method of the present invention is more uniform than that processed by the SIFT algorithm. In the area where the distribution of feature points is sparse, the position of the feature points extracted by the SIFT algorithm and the method in this paper are basically the sa...

Embodiment 2

[0043] Embodiment 2, as shown in Figure 3 (a), utilizes SIFT algorithm and the improved method of the present invention to match respectively, as shown in Figure 3 (b) and Figure 3 (c), the number of matching points of SIFT algorithm is 305 , the number of false matching point pairs is 18, and the correct matching rate is 94.10%; the number of matching point pairs of the improved algorithm of the present invention is 95, the number of false matching point pairs is 1, and the correct matching rate is 98.95%. It can be seen that the correct matching rate of the improved method of the present invention has increased by 4.85%.

Embodiment 3

[0044] Embodiment 3, as shown in Figure 4 (a), utilizes SIFT algorithm and the improved method of the present invention to match respectively, as shown in Figure 4 (b) and Figure 4 (c), the number of matching points of SIFT algorithm is 174 , the number of wrong matching point pairs is 17, and the correct matching rate is 90.23%; the number of matching point pairs in the improved method of the present invention is 72, the number of wrong matching point pairs is 2, and the correct matching rate is 97.22%. It can be seen that the correct matching rate of the improved method of the present invention has increased by 6.99%.

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 a PCA-SIFT-based fast image splicing method. The method includes the following steps that: an improved non-maximum value suppression method is introduced in a spatial extreme point detection stage to optimize initial feature points, so that a more uniformly distributed feature point set can be obtained; 64-dimensional SIFT descriptors are extracted based on a circular domain in a descriptor construction stage, a principal component analysis (PCA) method is adopted to perform further dimensionality reduction on the descriptors; a K-D tree BBF search strategy is introduced in a feature matching stage, and a random sampling consensus (RANSAC) algorithm is adopted to eliminate mismatched points, and therefore, matching speed and matching accuracy can be improved. An image splicing experiment shows that the correct matching rate of the method of the present invention is higher than that of an SIFT algorithm, and the splicing speed of the method of the present invention is 1.6 to 2.2 times that of a conventional SIFT algorithm.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a fast image stitching method based on PCA-SIFT. Background technique [0002] Image stitching refers to the stitching of several overlapping images into a wide-field, seamless high-resolution image. At present, image stitching technology is widely used in medical image diagnosis, military, remote sensing image processing and other fields. One of the most important steps in image stitching is image registration, and the quality of the image registration algorithm directly determines the quality and efficiency of image stitching. Mikolajczyk evaluated the performance of several representative image registration algorithms in terms of scale scaling, image compression, and perspective transformation. The results show that SIFT is the most effective image registration algorithm in the field of image processing at present, with strong Stickiness. However, this algorithm do...

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): G06T3/40G06K9/62G06K9/46
CPCG06T3/4038G06V10/462G06F18/2135
Inventor 郑茜颖杨炳坤程树英
Owner FUZHOU 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

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