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

Visual heading recognition method based on target point recognition

A recognition method and target point technology, applied in character and pattern recognition, image analysis, instruments, etc., can solve the problems of unguaranteed system reliability and cost increase, achieve simple static calibration, improve accuracy, and reduce calculation amount Effect

Inactive Publication Date: 2019-06-25
BEIHANG UNIV
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although the use of high-performance inertial components and processors can guarantee high reliability, the cost will increase greatly; the use of low-performance inertial components or low-cost visual sensors alone can reduce costs, but cannot guarantee system reliability

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
  • Visual heading recognition method based on target point recognition
  • Visual heading recognition method based on target point recognition
  • Visual heading recognition method based on target point recognition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0087] The present invention is a method for solving visual heading based on target point recognition. The method proposes a feature-based target point recognition algorithm. According to this algorithm and the camera calibration result, the visual heading can be calculated accurately and effectively. The implementation process of the method is as follows figure 1 As shown, the specific steps are as follows:

[0088] Step 1: According to figure 2 The imaging principle of the linear camera is shown, and the low-cost linear camera is simply and statically calibrated using the checkerboard to obtain the azimuth and pixel correspondence table, such as image 3 Shown is the image of the checkerboard target in the field of view;

[0089] Step 2: Load the azimuth and pixel correspondence table to obtain the first frame of image I(1), select the initial target point in I(1), such as Figure 5 As shown in (a), extract and describe ROI 1 (1) SURF feature points, store I(1) as the first refer...

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 visual heading recognition method based on target point recognition, which includes: loading the azimuth and pixel correspondence table obtained through camera calibration, selecting the initial target point in the first frame image, extracting and describing the first and SURF feature points in the second region of interest, and store this frame image as the first reference image; in the next frame of image collected, adjust the position and size of the region of interest, match the feature points with the reference image, and filter the ideal Coplanar feature points, calculate the homography matrix, and identify the current position of the target point; when the reference image update condition is met, update the current frame image as the reference image, and filter the stable ones that conform to the affine transformation and are located on the infinite plane The ideal coplanar feature point set is used to correct the homography array and the target point position; finally, the corresponding heading of the target point is identified according to the azimuth angle and pixel correspondence table.

Description

Technical field [0001] The invention relates to a method for visual heading recognition based on target point recognition, and belongs to the technical field of visual / inertial integrated navigation. Background technique [0002] The navigation system plays an important role in the complex motion process of autonomous mobile robots. The integrated navigation system, which is a combination of basic navigation technologies such as vision, inertial, GPS and GPS-like systems, ranging and direction recognition, has been extensively researched and developed. [0003] Visual navigation and inertial navigation are typical independent navigation systems. They can still provide reliable navigation services without the help of external conditions, especially when GPS is unavailable and other complicated situations. The real-time positioning and mapping (SLAM) algorithm is currently the most commonly used algorithm to realize the positioning and navigation of autonomous mobile robots. Based ...

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): G06K9/62G06T3/00G06T7/33
CPCG06T3/0075G06V10/751
Inventor 张海崔显光郭雷张晓鸥
Owner BEIHANG UNIV