To enhance the recovery capability of the polarization defogging algorithm for foggy scenes, a polarization image defogging algorithm founded upon the dark channel principle and multi-scale fusion is put forth. Initially, the atmospheric light is estimated based on the polarization information. This is followed by the estimation of atmospheric transmittance, which is achieved by combining the dark channel principle with the multi-scale image fusion technique. Finally, the atmospheric scattering model is constructed to obtain the defogged image. The experimental results demonstrate that the proposed algorithm exhibits excellent performance, achieving a high correlation coefficient of relevance and low CIEDE2000 color difference values in simulated fog environments of varying concentrations. The defogged image is improved by 59.68 times in terms of contrast, and the information entropy is increased by 11.38%. In a real outdoor fog scene, the defogged image generated by this algorithm showed a contrast improvement of 25.24 times and an information entropy enhancement of 8.32%. The presented algorithm defogs images more effectively and naturally than existing algorithms, enabling the recovery of detailed information from fog-obscured scenes.

Wavefront coding technology disperses incident laser energy by employing a pupil-phase mask, effectively reducing the energy density on the detector target and enhancing the camera’s anti-laser capability. Typically, phase masks with large modulation coefficients are used to achieve better protection performance, but this introduces significant challenges for subsequent digital restoration. The digital restoration process is a classical ill-posed problem, requiring simultaneous blur removal and noise suppression. Traditional image restoration methods often suffer from high-frequency noise amplification and processing artifacts. To address the challenge of restoring complex degraded images after encoding, a ResUNet model incorporating global and local attention mechanisms is constructed. This model synergistically extracts global contextual information and local detailed features of the image, overcoming limitations of traditional linear decoding methods, such as a strong dependence on noise priors and susceptibility to reconstruction artifacts. Simulations and experimental validations demonstrate that the proposed decoding algorithm effectively restores high-frequency image details and significantly improves reconstruction quality, outperforming conventional methods. Tests on real captured images at different distances show that the network model achieves optimal performance across multiple evaluation metrics, with an average PSNR of 21.75 dB, confirming its excellent image restoration capability. This research advances the practical application of wavefront encoding technology in laser protection, enabling systems to maintain high image quality while providing laser protection capabilities.

At detection ranges on the order of kilometers, cat-eye echoes degrade into pixel-level spots and exhibit responses similar to those of compact, high-echo false targets, making it difficult for single-band systems to balance weak-target detection rates with false alarm control. To address this, this paper pairs spatially registered and radiometrically calibrated 808 nm and 905 nm images to form a dual-band detection pair, utilizing the cross-band response asymmetry caused by chromatic defocus for discrimination. Experiments show that cat-eye targets exhibit coupled asymmetry in intensity, scale, and energy distribution, whereas the false targets under test generally maintain approximate symmetry; furthermore, the dominant discriminative information shifts from scale differences to intensity differences as distance increases. Based on this pattern, this paper proposes a range-conditioned “detection–reclassification” framework that dynamically matches appropriate models based on target distance. In this framework, a structurally symmetric, two-branch, multi-scale attention-fused YOLO network is used for candidate target detection, while an explicit asymmetric feature classifier further processes challenging samples to retain weak targets and suppress false alarms. Field experiments conducted at distances ranging from 100 to 2100 m achieved an overall mAP50 of 0.907 and an overall mAP50–95 of 0.475, with the mAP50 improving by 17.7 and 10.8 percentage points, respectively, compared to the 808 nm and 905 nm single-band baselines. The results demonstrate that this framework can effectively balance the detection of weak targets at long ranges with the suppression of false targets under the evaluated conditions.