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

Weather image recognition method based on CNN and multi-feature fusion

A multi-feature fusion and weather image technology, applied in the field of image processing, can solve the problem of weak acquisition of sensitive clues, achieve high reliability and improve accuracy

Active Publication Date: 2019-12-10
CHENGDU UNIV OF INFORMATION TECH
View PDF9 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

CNN can obtain most of the basic information of the image, but the ability to obtain sensitive clues that can represent weather phenomena is weak
There is currently no recognition method that combines CNN features and weather features

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
  • Weather image recognition method based on CNN and multi-feature fusion
  • Weather image recognition method based on CNN and multi-feature fusion
  • Weather image recognition method based on CNN and multi-feature fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0059] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0060] 1. Extract six weather features that can represent different weather phenomena and encode them into feature vectors. The weather features include the brightness value in the weather image, the difference between the maximum and minimum pixel intensity in the image, that is, the contrast value, and the weather image. The fog factor, image sharpness values, white pixel values ​​in the image, and the image's color histogram.

[0061] 1. Brightness value Y'

[0062] Brightness explains weather images well and is one of the most important pixel properties. For example, sunny images usually have higher brightness, while cloudy and hazy images usually have lower brightness. Encode the brightness information into a feature vector: Y'=0.299r+0.587g+0.114b, where: r, g, and b represent the pixel values ​​of each pixel in the image in 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 weather image recognition method based on CNN and multi-feature fusion, and the method comprises the following steps: extracting six weather features representing different weather phenomena for an input image, and coding the weather features into a feature vector; extracting high-dimensional CNN features of the representation image; performing feature fusion on the weather feature vector and the CNN feature vector to form an overall feature vector; and training a classification model by using the overall feature vector, and identifying the weather image by using thetrained classification model. The weather features and the CNN features are fused for training and classification, and the recognition accuracy is high.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a weather image recognition method based on CNN and multi-feature fusion. Background technique [0002] Severe weather conditions often lead to catastrophic events such as ship collisions, forest fires, power plant grid failures, train derailments and plane crashes. Many scholars have already done some work on exploring and solving such problems of weather recognition: traditional weather recognition methods rely on special sensors (special microwave sensors or satellite imaging sensors) and human eyes as an aid for recognition, however This method consumes a lot of manpower and material resources, and this semi-automatic recognition accuracy is low. Recently, Convolutional Neural Network (CNN) has made a major breakthrough in the field of computer vision, especially image recognition, and the research on automatic recognition of weather phenomena based on CNN has received extens...

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): G06K9/62
CPCG06F18/24765G06F18/2411G06F18/253
Inventor 李英祥李志强任堃钟剑丹
Owner CHENGDU UNIV OF INFORMATION TECH