Camouflaged Target Detection Based on Snapshot Multispectral Imaging

The spectral information contained in the hyperspectral images (HSI) distinguishes intrinsic properties of a target from background, which is widely used remote sensing. However, low imaging speed and high data redundancy caused by resolution spectrometers limit their application scenarios with real-time requirement. In this work, we achieve precise detection camouflaged targets based on snapshot multispectral technology band selection methods urban-related scenes. Specifically, algorithm combines constrained energy minimization (CEM) improved maximum between-class variance (OTSU) (t-OTSU), proposed to obtain initial results adaptively segment region. Moreover, an object region extraction (ORE) complete contour that improves capability (MSI). experimental show has ability detect different using only four bands. accuracy above 99%, false alarm rate below 0.2%. research achieves effective potential provide new means for sensing complex