A wavelength-dependent extinction coefficient model for multispectral image defogging

Most of the state-of-the-art defogging models presented in the literature assume that the attenuation coefficient of all spectral channels is a constant, which inevitably leads to spectral distortion and information bias. To address this issue, this paper proposes a novel defogging method that takes into account the difference between the extinction coefficients of multispectral channels of light traveling through fog. Then the spatially distributed transmission map of each spectral channel is reconstructed to restore the fog-degraded images. The experimental results of various realistic complex scenes show that the proposed method has outstanding advantages in restoring lost detail, compensating for degraded spectral information, and recognizing targets hidden in dense fog. In addition, this work provides a rare method to characterize the intrinsic property of fog expressed as multispectral relative extinction coefficients, which act as a fundament for further reconstruction of multispectral information.