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

Synchronous positioning and mapping method using residual attention mechanism network

A technology of synchronous positioning and attention, applied in neural learning methods, biological neural network models, image analysis, etc., can solve the problem of high redundancy of geometric feature information, and achieve the effect of easy understanding and solving six-degree-of-freedom pose problems

Active Publication Date: 2020-05-05
TONGJI UNIV
View PDF2 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The larger the extent, the map will grow proportionally, and this is because the redundancy of geometric feature information in traditional methods is too high

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
  • Synchronous positioning and mapping method using residual attention mechanism network
  • Synchronous positioning and mapping method using residual attention mechanism network
  • Synchronous positioning and mapping method using residual attention mechanism network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0095] In this embodiment, the image is first input into the attention mechanism network. In the case of cluttered background, different types of attention are needed to simulate images with complex scenes and large appearance changes. In this case, features from different layers need to be modeled by different attention masks. The incremental nature of the stacked network structure can gradually increase attention to complex images. The trunk branch performs feature processing. The subsequent LSTM module ensures that the attention distribution in the image is relevant to the location prediction. To be able to find and exploit correlations between images taken in long trajectories, long-short-term memory gates capable of learning long-term dependencies by introducing memory gates and cells are used as subsequent network structures. Correspondingly, although LSTM gates can handle long-term dependencies and have a deep temporal structure, it still requires depth on network lay...

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 relates to a synchronous positioning and mapping method using a residual attention mechanism network. The method comprises the following steps: step 1, training a neural network; step 2,inputting a group of pictures into a neural network to obtain an RGB image and a feature weight image corresponding to each picture; step 3, performing improved FAST corner detection on the RGB image; step 4, selecting final feature points; step 5, matching the feature points, and solving camera initialization motion for polar constraint; step 6, solving local camera motion; step 7, performing loop detection between images and acquiring an accurate track of a camera; and step 8, carrying out dense reconstruction to obtain an environment map. Compared with the prior art, the method has the advantages that the feature points are easier to understand by people, important areas can be highlighted through colors and brightness, and the like.

Description

technical field [0001] The invention relates to a method for synchronous positioning and mapping, in particular to a method for synchronous positioning and mapping using a residual attention mechanism network. Background technique [0002] Simultaneous positioning and mapping is a classic problem in the computer field. It has been widely studied in the fields of image processing and computer vision, but it is still a challenging problem. It refers to the process in which a moving object calculates its own position and builds an environmental map based on the information of the sensor. Traditional simultaneous localization and mapping methods are mainly based on feature point method to estimate camera motion. Because the picture itself contains too much information, most methods select representative points from the image, which are called landmarks in the classic synchronous positioning and mapping method. [0003] Another potentially more promising research direction for ...

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
IPC IPC(8): G06T7/246G06T7/73G06T7/80G06N3/04G06N3/08
CPCG06T7/246G06T7/80G06N3/08G06T7/73G06N3/044Y02T10/40
Inventor 张佳伟尤鸣宇
Owner TONGJI UNIV