3D Range-gated Imaging Research Articles

Depth-prior-based range-gated 3D imaging method: integration of optical property into visual learning

3D Range-gated Imaging (3DRGI) has great potential for long-range detection in adverse weather conditions. Recently, vision-guided 3DRGI has brought new perspectives to this area as it overcomes hardware limitations and greatly increases flexibility. However, existing vision-guided methods do not consider the optical properties of range-gated imaging, which results in low accuracy. This paper proposes a depth-prior-based 3DRGI method to combine the advantages of optical-based and vision-guided methods. In this method, depth-prior is firstly deduced from principles of range-gated imaging and provides effective depth signals. Then, adaptive depth intervals are estimated using statistical methods, and a depth-prior-guided loss function is designed. The integration of the depth-prior-guided loss function within a vision-guided model enables focused attention on pixels with depth estimations that are inconsistent with the depth prior, thereby refining the overall accuracy of the depth map. To prove the feasibility of the proposed method, comparison experiments and ablation studies have been performed. The results show that the root mean square error improved by more than 12% under adverse weather conditions.

Bayesian reconstruction method for underwater 3D range-gated imaging enhancement.

We investigate a systematic improvement for 3D range-gated imaging in scattering environments. Drawbacks including absorption, ambient light, and scattering effect are studied. The former two are compensated through parameter estimation and preprocessing. With regard to the scattering effect, we propose a new 3D reconfiguration algorithm using a Bayesian approach that incorporates spatial constraints through a general Gaussian Markov random field. The model takes both scene depth and albedo into account, which provides a more informative and accurate restoration result. Hyper-parameters for the statistical mechanism are evaluated adaptively in the procedure and an iterated conditional mode optimization algorithm is employed to find an optimum solution. The performance of our method was assessed via conducting various experiments and the results also indicate that the proposed method is helpful for restoring the 2D image of a scene with improved visibility.

水下二维及三维距离选通成像去噪技术研究

Compared with traditional underwater cameras, the detection distance of underwater range-gated imaging can be increased by 2-3 times. Furthermore, based on this technology, fast high-resolution 3D imaging can be achieved. It has great potentials in underwater target detection and recognition, automatic navigation, marine scientific research and natural resources exploration. Although range-gated imaging can suppress the backscattering noise by space slicing, and achieve higher quality images, the backscattering noise of the sampling water volume still exists in gated images, resulting in low image signal-to-noise ratio and contrast, especially for distant targets or low reflectance targets. This paper systematically introduced author's research of deblurring methods for 2D and 3D range-gated imaging. In 2D imaging, two methods were proposed:the first was an algorithm of self-adaptive double-plateau histogram equalization to improve image contrast and better meet human vision; the second was to use a water-noise-reference denoising algorithm to improve the signal-to-noise ratio of target gated images. In 3D imaging, there were also two methods proposed:one was to use the water-noise-reference denoising algorithm for 3D reconstruction, and the other was to use the threshold-dependent joint bilateral filter algorithm for enhancing 3D images. The four proposed methods can be used independently or in combination for denoising enhancement of underwater range gated imaging.

An Application of Homogeneity Filter to Complete Image Details in 3D Images

In order to complete details in the method of 3D range-gated imaging based on the range-intensity profiles (RIP), we apply homomorphic filtering to tomographic-like serial images in succession. This application filters noise disturbance from low laser illuminating homogeneity, atmospheric scattering and ambient background. First, tomographic-like serial images are filtered by EHPF, BHPF and GHPF, respectively, and three 3D images are reconstructed by RIP correlation algorithm at the same time. And then compare three 3D images under homomorphic filtering and 3D image reconstructed by raw tomographic-like serial images. At last, we analyze and evaluate gray levels in specific regions, which are extracted from tomographic-like serial images and 3D images, correspondingly. The experiments results show that homomorphic filtering is effective in noise elimination and improvement in completing details in 3D images.

Adaptive Polarization-Modulated Method for High-Resolution 3D Imaging

To improve the resolution in 3D range-gated imaging, an adaptive polarization-modulated method using electron multiplying charge-coupled devices (EMCCDs) is proposed in this letter. With polarization beam splitting, the returned light is separated into two orthogonal polarized components, which are modulated by cos2 function and sin2 function, respectively. Based on a dual EMCCDs structure, both a depth map and an intensity image can be reconstructed from the two modulated images instantaneously, which will enhance the performance of dynamic imaging. The results show that the range accuracy of the interested objects in the scene can be promoted from 4.4 to 0.26 m with the art of adaptive gate-opening range.

Triangular-range-intensity profile spatial-correlation method for 3D super-resolution range-gated imaging

We present a triangular-range-intensity profile (RIP) spatial-correlation method for 3D range-gated imaging with a depth of super resolution. In this method, spatial sampling volumes with triangular-RIPs are established by matching laser-pulse width and sensor gate time, and then depth information collapsed in gate images can be reconstructed by spatial correlation of overlapped gate images corresponding to sampling volumes. Compared with super-resolution depth mapping under trapezoidal-RIPs, range accuracy and precision are improved, and a large range fluctuation due to noise disturbance is smoothed by noise suppression under triangular-RIPs. In this paper, a proof experiment is demonstrated with a range precision 2.5 times better than that obtained under trapezoidal-RIPs.