研究目的
To improve the accuracy of overheating fault localization in electric equipment monitoring systems by fusing infrared and visible images using a novel method based on Finite Discrete Shearlet Transform (FDST) and Dual-Channel Pulse Coupled Neuron Network (DC-PCNN).
研究成果
The combination of FDST and DC-PCNN provides a more efficient way to integrate complementary information from infrared and visible images, significantly improving the accuracy of overheating fault localization in electric equipment monitoring systems.
研究不足
The proposed method may require longer computational time compared to simpler fusion rules, and the effectiveness of some objective evaluation criteria can be confusing or deceptive.
1：Experimental Design and Method Selection:
The proposed method uses FDST for image decomposition and DC-PCNN for subband fusion with different linking strengths for low-frequency and high-frequency subbands.
2：Sample Selection and Data Sources:
The experiments are conducted on images from the TNO multiband image data collection and actual substation electric equipment under working conditions.
3：List of Experimental Equipment and Materials:
MATLAB 2016b on an i3-7100
4：90GHz PC with 0GB RAM. Experimental Procedures and Operational Workflow:
Decompose source images via FDST, calculate modified spatial frequency (MSF) in subband coefficients, fuse subbands using DC-PCNNs, and reconstruct the fused image by inverse FDST.
5：Data Analysis Methods:
Mutual information (MI), QAB/F, and standard deviation (SD) are used as objective evaluation criteria.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
