No abstract

No abstract

In underwater range-gated imaging (URGI), enhancement of low-brightness and low-contrast images is critical for human observation. Traditional histogram equalizations over-enhance images, with the result of details being lost. To compress over-enhancement, a lower-upper-threshold correlation method is proposed for underwater range-gated imaging self-adaptive enhancement based on double-plateau histogram equalization. The lower threshold determines image details and compresses over-enhancement. It is correlated with the upper threshold. First, the upper threshold is updated by searching for the local maximum in real time, and then the lower threshold is calculated by the upper threshold and the number of nonzero units selected from a filtered histogram. With this method, the backgrounds of underwater images are constrained with enhanced details. Finally, the proof experiments are performed. Peak signal-to-noise-ratio, variance, contrast, and human visual properties are used to evaluate the objective quality of the global and regions of interest images. The evaluation results demonstrate that the proposed method adaptively selects the proper upper and lower thresholds under different conditions. The proposed method contributes to URGI with effective image enhancement for human eyes.

For the in situ detection technologies of planktons and fishes, optical cameras traditionally have a small and fixed sampling volume with a strong target-sized dependent (typically<1  mm), and imaging sonar has lower spatial resolution (typically>2  cm) with a problem of species identification. To solve the above problems, this paper proposes an in situ detection method of optical gated sampling for millimeter- to centimeter-scale plankton and fish detection. In this method, the sampling volume can be flexibly adjusted by matching the temporal parameters of gate pulses and illuminator laser pulses to satisfy target observation with different sizes. The gated sampling suppresses the backscattering of water and also filters the environment background so that transparent planktons can be detected by high contrast. Furthermore, the sampling volume is determined by the convolution of gate pulses and laser pulses, and thus the target abundance is derived. Theory and simulation of abundance measurement are established. In experiments, transparent jellyfishes are recorded with a spatial resolution of better than 100 μm. In addition, proof experiments of sampling volume adjustment and abundance measurement are demonstrated.