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

A Multi-domain Data Cleaning and Learning Method Based on Coupling Modeling

A technology of data cleaning and learning method, applied in the field of pattern recognition, which can solve the problems of depth data estimation error, reducing scene learning and detection accuracy, etc.

Active Publication Date: 2021-02-05
HOHAI UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the dark channel estimation model is extremely sensitive to sporadic noise points, which leads to serious errors in depth data estimation and reduces the accuracy of scene learning and detection.

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
  • A Multi-domain Data Cleaning and Learning Method Based on Coupling Modeling
  • A Multi-domain Data Cleaning and Learning Method Based on Coupling Modeling
  • A Multi-domain Data Cleaning and Learning Method Based on Coupling Modeling

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0021] Below in conjunction with specific embodiment, further illustrate the present invention, should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention, after having read the present invention, those skilled in the art will understand various equivalent forms of the present invention All modifications fall within the scope defined by the appended claims of the present application.

[0022] The overall flow of the method disclosed in the present invention, such as figure 1 As shown, the typical examples of scattering environment are dense fog weather and underwater environment with visibility less than 50m. In these two examples, due to the scattering and refraction of light by suspended particles in air and water, a typical scattering environment is formed. Through the calculation of the original scene imaging data in the scattering environment, the color field data and the...

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 multi-field data cleaning and learning method based on coupling modeling, which extracts multi-field data and realizes the identification and detection of scene content through the data cleaning and learning method. Collect and calculate multi-field data in the same scene; use coupling modeling between multi-field data for mutual verification to realize data cleaning; use domain adaptive learning mechanism to learn and integrate multi-field data under the same framework to form a unified The multi-domain joint scene model under the framework can detect scene objects. The invention can be stably and reliably used for knowledge learning and target detection under complex conditions, can fully mine scene information, has better noise suppression effect and higher computing efficiency.

Description

technical field [0001] The invention relates to a pattern recognition method, in particular to a coupling modeling-based multi-field data cleaning and learning method for target detection. Background technique [0002] Pattern recognition and target detection methods in high scattering and strong attenuation environments have been long-term problems in related fields. First of all, the scattering environment will cause serious attenuation of target information, making it difficult to accurately identify the difference between the target and the background. Secondly, the scattering environment will interfere with the apparent characteristics of the target, and the target features will be seriously distorted and mixed with false target information. In order to overcome this problem, existing methods mostly use preprocessing methods to suppress environmental noise and enhance target features. However, extensive research and testing has found that this preprocessing-dependent ...

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): G06F16/215G06K9/00G06T7/50G06T7/90
Inventor 陈哲李臣明石爱业徐立中
Owner HOHAI UNIV