Single-photon compression recognition system based on binary neural network and using method thereof

A binary neural and identification system technology, applied in the field of single-photon compression identification system, can solve the problems of low identification accuracy, complex scheme, long time-consuming, etc., and achieve the effect of high identification accuracy and fast speed.

Inactive Publication Date: 2021-02-09
NANCHANG UNIV
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the recognition of this type of method requires image reconstruction of the target first, so there are disadvantages such as complex scheme, long time consumption and low recognition accuracy at extremely low sampling rate

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
  • Single-photon compression recognition system based on binary neural network and using method thereof
  • Single-photon compression recognition system based on binary neural network and using method thereof
  • Single-photon compression recognition system based on binary neural network and using method thereof

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0029] The present invention will be further described below in conjunction with drawings and embodiments.

[0030] Such as figure 1 Shown is the neural network (BFCI-Net) of the present invention (the single-photon compression recognition system based on binary neural network), which reshape the picture into a column of vectors, and then input it to the binary neural network for training. The first layer is a binary fully connected layer with weight matrix size (x, y), the size of x depends on the size of the picture, and the size of y depends on the sampling rate MR of compressed sensing, that is, y=MR*x , the weight matrix obtained through training is the measurement matrix of compressed sensing. After several layers of fully connected layers and batch normalization layers, the prediction results of the pictures are output from the output layer.

[0031] 1.1 Compressed sampling subnetwork F c (·)

[0032] The first fully connected layer in the neural network (BFCI-Net) ...

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 belongs to the technical field of single photon imaging, and particularly relates to a single-photon compression identification system based on a binary neural network and a using methodthereof, the single photon compression identification system comprises a binary compression sampling sub-network and an identification sub-network, a weight matrix of the binary compression samplingsub-network is a binary matrix composed of + 1 and 1, and the binary compression sampling sub-network is a binary matrix composed of + 1 and 1. The '+ 1' and the '1' are respectively used for modulating forward turning and reverse turning of a DMD micromirror in the single-pixel imaging system; the identification sub-network is used for carrying out dimension compression and expansion on the measurement value matrix and completing category identification to obtain an identification result, and the weight of the identification sub-network is obtained by network training. According to the invention, the identification speed is faster; the invention can complete the recognition of images without image reconstruction, is higher in recognition accuracy, and can maintain the higher recognition accuracy under the condition of extremely low sampling rate.

Description

technical field [0001] The invention belongs to the technical field of single-photon imaging, and in particular relates to a binary neural network-based single-photon compression recognition system and an application method thereof. Background technique [0002] Single-pixel imaging is an imaging method that uses a single-point detector to capture a two-dimensional image. In 2008, Romber and Baraniuk of Rice University proposed a single-pixel camera scheme based on the principle of compressed sensing. This scheme uses Spatial sparsity of images and point detectors enable imaging of 2D objects. In 2012, Yu Wenkai of the Chinese Academy of Sciences combined single-pixel photography technology with photon counting imaging technology to realize ultra-high-sensitivity photon counting imaging that broke through the limit of the detector itself. This imaging solution has the advantages of high sensitivity and low photon level. [0003] With the development of computer vision techn...

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): G06K9/00G06K9/42G06N3/04G06N3/08G02B26/08
CPCG06N3/04G06N3/08G02B26/0833G06V30/32G06V10/32G06F18/00
Inventor 鄢秋荣祝志太熊乙宁蔡源鹏杨耀铭
Owner NANCHANG UNIV
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