Application of Deep Boltzmann Machine to Image Fusion

Journal Title: 河南科技大学学报(自然科学版) - Year 2018, Vol 39, Issue 1

Abstract

Aiming at the problem of insufficient feature information extraction during the fusion course of visible light and infrared images, the fusion method for visible light and infrared images based on non-subsampled shearlet transform (NSST) and deep Boltzmann machine (DBM) was proposed.First, optimal energy segmentation on the infrared image was conducted via DBM to extract the significant infrared target.Secondly, the regions of significant targets and backgrounds via segmentations were projected to the source images respectively to obtain corresponding maps.Finally, the multi-scale and multi-direction sparse decompositions of visible light and infrared projection images were performed by NSST, respectively.A series of sub-images of the source images were fused.The final fused image can be obtained by inverse NSST.The results of the simulation experiment show that compared with three common methods, the method discussed in the paper is higher than contourlet transform (CT) 4.1% in information entropy.It's better than NSST about 10% in average gradient, and it is higher than non-subsampled contourlet transform (NSCT) about 3.8% in mutual information.The fusion images acquired from the method discussed in the paper has a higher contrast degree and more specific information.

Authors and Affiliations

Fengnian LIU

Keywords

Related Articles

Indoor Fingerprint Localization Method Using Channel State Information

Received signal strength indicator ( RSSI) is always used to generate fingerprint in conventional indoor fingerprint positioning technology. However, RSSI is vulnerable to environmental changing, and the positioning accu...

Vehicle Real Driving Cycle Construction Based on Improved Fuzzy C-means Clustering Algorithm

Since the performance of Fuzzy C-means ( FCM) clustering algorithm depended on the initial clustering center, the iterative was inclined to fall into the local extremum and could not ensure that FCM converges to an optim...

Development of Small Wind External Rotor Permanent-magnet Generator Experimental Platform

Small wind external rotor permanent magnet generator( PMG) experimental platform was developed for testing its performance. The experimental platform was mainly composed of hardware unit and software unit.The hardware wa...

Nonlinear Dynamic Response Analysis of Rotor-Ball Bearing System For Momentum Wheel

The dynamic equation of the rotor-bearing system of momentum wheel considering the nonlinear bearing force was established and solved by Runge-Kutta numerical method. The bifurcation diagrams, Poincare maps and amplitude...

Design of Alternating Current Charging Spot Control System for Electric Vehicle

By analyzing the system structure and working principles of alternative current( AC) charging spot,the control system of AC charging spot for electric vehicle was designed based on STM32 processor. The basic functions an...

Download PDF file
  • EP ID EP464632
  • DOI -
  • Views 73
  • Downloads 0

How To Cite

Fengnian LIU (2018). Application of Deep Boltzmann Machine to Image Fusion. 河南科技大学学报(自然科学版), 39(1), -. https://europub.co.uk/articles/-A-464632