Real-time detection method of visible light remote-sensing image cloud field

A technology for remote sensing images and detection methods, which is applied in the fields of instruments, scene recognition, and computing, and can solve problems such as high computational complexity and unsatisfactory real-time cloud area detection.

Inactive Publication Date: 2015-09-23
HUAZHONG UNIV OF SCI & TECH +1
View PDF2 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the computational complexity of these texture features is high, which does not meet the requirements of real-time cloud region 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
  • Real-time detection method of visible light remote-sensing image cloud field
  • Real-time detection method of visible light remote-sensing image cloud field
  • Real-time detection method of visible light remote-sensing image cloud field

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0053] Embodiment 1: Example of detection of cloud areas containing large cloud areas

[0054] 1. Basic processing unit division

[0055] In order to control the processing granularity of the cloud area detection algorithm, the test image is first divided into non-overlapping G×G image sub-blocks. In this experiment, the test image is a panchromatic image, the pixel data bit depth is 8 bits, the spatial resolution is 2 meters, and the image block size G is set to 64.

[0056] 2. Whitening treatment

[0057] In order to maintain the invariance of feature extraction to illumination changes, the image sub-block is whitened, that is, all pixels in the image sub-block minus the mean value of the pixel value in the sub-block, divided by the standard deviation of the pixel value in the sub-block. The result of the whitening process is used as the input of the filter.

[0058] 3. Filtering based on unidirectional anisotropic Gaussian filter bank

[0059] Since the appearance of cl...

Embodiment 2

[0069] Embodiment 2: Example of detection of cloud areas containing small cloud areas

[0070] Its processing step is identical with embodiment 1. Figure 3(a) is the original remote sensing image containing a small cloud area, Figure 3(b) is the detection result of the cloud area based on the histogram feature of the visual word in six directions without dimensionality reduction, and Figure 3(c) is the detection result based on the single direction and The cloud area detection result of the visual word histogram feature without dimensionality reduction, Fig. 3 (d) is the cloud area detection result of the visual word histogram feature proposed by the present invention. By comparing Fig. 3(b), Fig. 3(c) and Fig. 3(d), the visual word histogram feature proposed by the present invention can achieve detection performance equivalent to that of the two variant feature extraction algorithms.

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 real-time detection method of a visible light remote-sensing image cloud field, and belongs to an image data processing method. The real-time detection method comprises the following steps: carrying out original image subblock division, then, quickly extracting the characteristics of each subblock, and finally, finishing cloud field detection through the dichotomy of the characteristics of each subblock. The invention provides a method that a filtering operation is implemented by a unidirectional anisotropy Gaussian filter group by aiming at the characteristics of the no direction of the cloud field. In order to avoid the complexity of a large-scale two-dimensional convolution operation, a unidirectional anisotropy Gaussian filtering operation is realized through an infinite impulse response filter of row and column separation. In order to lower visual word query complexity in a generation process of visual word histogram characteristics, a visual word is subjected to characteristic dimensionality reduction through main ingredient analysis, then, query is carried out, query precision can be guaranteed by a dimensionality reduction query strategy, and query complexity can be greatly lowered. The method can meet the requirements of on-satellite real-time cloud field detection.

Description

technical field [0001] The invention belongs to image data processing methods, in particular to a real-time visible light remote sensing image cloud area detection method. Background technique [0002] Cloud area detection is one of the primary tasks and key technologies of visible light remote sensing image compression, transmission and processing. For a long time, visible light remote sensing images have been widely used due to their high resolution and high definition. However, an obvious weakness of visible light images is that cloud occlusion significantly reduces the observability of ground target information, and some thick cloud occlusion images are even unusable. Real-time thick cloud coverage area detection can play a positive role in improving the compression efficiency of remote sensing images, saving data transmission bandwidth and improving processing efficiency. From the data processing flow point of view, the cloud detection algorithm is at the forefront of...

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): G06K9/00G06K9/62
CPCG06V20/13G06V20/194G06F18/23213G06F18/2411
Inventor 李彦胜李情芸任菲菲刘磊伊成俊杜辉余毓杰谭毅华闫雪梅田金文
Owner HUAZHONG UNIV OF SCI & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap