Denoising and Feature Extraction for Space Infrared Dim Target Recognition Utilizing Optimal VMD and Dual-Band Thermometry

Space target feature extraction and space infrared recognition are important components of situational awareness (SSA). However, owing to far imaging distance between the detector, signal received by detector is dim easily contaminated noise. To effectively improve accuracy recognition, it essential suppress noise signal. Hence, a novel denoising extracting method combinating optimal variational mode decomposition (VMD) dual-band thermometry (DBT) proposed. It takes mean weighted fuzzy-distribution entropy (FuzzDistEn) band-limited intrinsic functions (BLIMFs) as optimization index dragonfly algorithm (DA) obtain parameters (K, α) VMD. Then VMD utilized decompose noisy series BLIMFs Pearson correlation coefficient (PCC) proposed determine effective modes reconstructe Finally, based on signal, temperature emissivity-area product calculated using DBT. The simulation experiment results show that has better reduction performance compared with other methods, improved at different equivalent irradiance. This provides more accurate temerpature for recognition.