Underwater image super-resolution using multi-stage information distillation networks

Recently, single image super-resolution (SISR) has been widely applied in the fields of underwater robot vision and obtained remarkable performance. However, most current methods generally suffered from the problem of a heavy burden on computational resources with large model sizes, which limited their real-world underwater robotic applications. In this paper, we introduce and tackle the super resolution (SR) problem for underwater robot vision and provide an efficient solution for near real-time applications. We present a novel lightweight multi-stage information distillation network, named MSIDN, for better balancing performance against applicability, which aggregates the local distilled features from different stages for more powerful feature representation. Moreover, a novel recursive residual feature distillation (RRFD) module is constructed to progressively extract useful features with a modest number of parameters in each stage. We also propose a channel interaction & distillation (CI&D) module that employs channel split operation on the preceding features to produce two-part features and utilizes the inter channel-wise interaction information between them to generate the distilled features, which can effectively extract the useful information of current stage without extra parameters. Besides, we present USR-2K dataset, a collection of over 1.6K samples for large-scale underwater image SR training, and a testset with an additional 400 samples for benchmark evaluation. Extensive experiments on several standard benchmark datasets show that the proposed MSIDN can provide state-of-the-art or even better performance in both quantitative and qualitative measurements.     

Sounds Stimulation on In Vitro HL1 Cells: A Pilot Study and a Theoretical Physical Model

Mechanical vibrations seem to affect the behaviour of different cell types and the functions of different organs. Pressure waves, including acoustic waves (sounds), could affect cytoskeletal molecules via coherent changes in their spatial organization and mechano-transduction signalling. We analyzed the sounds spectra and their fractal features. Cardiac muscle HL1 cells were exposed to different sounds, were stained for cytoskeletal markers (phalloidin, beta-actin, alpha-tubulin, alpha-actinin-1), and studied with multifractal analysis (using FracLac for ImageJ). A single cell was live-imaged and its dynamic contractility changes in response to each different sound were analysed (using Musclemotion for ImageJ). Different sound stimuli seem to influence the contractility and the spatial organization of HL1 cells, resulting in a different localization and fluorescence emission of cytoskeletal proteins. Since the cellular behaviour seems to correlate with the fractal structure of the sound used, we speculate that it can influence the cells by virtue of the different sound waves’ geometric properties that we have photographed and filmed. A theoretical physical model is proposed to explain our results, based on the coherent molecular dynamics. We stress the role of the systemic view in the understanding of the biological activity.     

基于GA-BP的汽车风振噪声声品质预测模型

目前对于汽车风振噪声的优化研究主要以声压级（Sound pressure level，SPL）作为单一评价指标，既不能全面反映噪声的物理属性，也无法考虑人耳对噪声的主观认知过程。为准确评价风振噪声，引入声品质，运用大涡模拟（Large eddy simulation，LES）对风振噪声进行数值仿真，根据实车道路试验判断仿真的准确性；对仿真结果进行声品质客观评价与主观评价，综合声品质客观评价参数与声品质主观评价试验结果建立BP神经网络预测模型；利用遗传算法（Genetic algorithm，GA），进一步对BP神经网络的结构参数进行优化，建立GA-BP声品质预测模型。研究结果表明，GA-BP声品质预测模型在训练速度和预测精度上都优于BP神经网络预测模型。预测模型基于声品质主客观评价结果，其预测值可以代替传统的声压级评价指标，为风振噪声提供更为准确合理的评价。     

A survey on tracking control of unmanned underwater vehicles: Experiments-based approach

This paper aims to provide a review of the conceptual design and theoretical framework of the main control schemes proposed in the literature for unmanned underwater vehicles (UUVs). Additionally, the objective of the paper is not only to present an overview of the recent control architectures validated on UUVs but also to give detailed experimental-based comparative studies of the proposed control schemes. To this end, the main control schemes, including proportional–integral–derivative (PID) based, sliding mode control (SMC) based, adaptive based, observation-based, model predictive control (MPC) based, combined control techniques, are revisited in order to consolidate the principal efforts made in the last two decades by the automatic control community in the field. Besides implementing some key tracking control schemes from the classification mentioned above on Leonard UUV, several real-time experimental scenarios are tested, under different operating conditions, to evaluate and compare the efficiency of the selected tracking control schemes. Furthermore, we point out potential investigation gaps and future research trends at the end of this survey.     

基于CFD⁃DEM的水下切粒装置水室内颗粒流动过程数值模拟

基于CFD?DEM耦合方法，研究了颗粒在水室内的流动状态，分析了不同刀盘转速、粒子水通入量和水室出口角度对造粒过程的影响，发现提高刀盘转速、增加粒子水通入量和水室出口倾斜一定的角度都有利于水室内颗粒的排出。进一步研究了颗粒与碎屑在水室内的流动，发现在水室出口处二者的流动基本呈现出一定的分离角度。     

自适应背景光估计与非局部先验的水下图像复原

有效地实现单幅水下降质图像复原对水下资源探索及环境监控领域的清晰图像获取具有极其重要的意义。为解决常用暗通道先验方法来复原图像时,背景光的估计易受白色物体干扰,且无法有效估计前景中白色物体透射率,复原质量不高的问题。本文提出了自适应背景光估计与非局部先验的水下图像复原算法。首先根据背景光具有高亮度及平坦性的特点,利用阈值分割算法获得背景光的候选区域,再通过图像的色调信息从候选信息中选取最佳的背景光点。随后,利用各颜色通道光的波长与散射系数的相关性,提出了适用于水下图像的非局部先验,并利用该先验估计各通道的透射率。最后针对复原结果中,因水下介质,微生物,水流影响而产生的加性噪声,设计去噪的最小优化问题,并利用引导滤波求解该问题,以去除复原结果中的加性噪声。实验表明:该算法在确保运行效率的基础上,准确地估计透射率,较常用算法的复原精度提高了约18%。证明了该算法能有效用于单幅水下图像复原的工程实践中。     

不锈钢网负载Al-beta分子筛涂层的制备及其在油水分离中的应用

使用二次生长法在不锈钢网表面制备了Al-beta分子筛涂层，并通过扫描电镜、X射线衍射和接触角检测对其进行了表征。Al-beta分子筛晶粒以球形呈现在不锈钢网表面，并相互交联而构成微纳双级粗糙结构，Al-beta涂层表现出超亲水和水中超疏油性质。将不锈钢网负载Al-beta分子筛涂层用于分离一系列油水混合物，并考察了其耐久性和稳定性。实验结果表明，不锈钢网负载Al-beta分子筛涂层在重复使用100次后其正己烷/水分离效率仍保持在97.1%以上，酸、碱和超声等处理对其形貌和油水分离效率几乎没有影响，具备优异的耐久性、自清洁性和稳定性，在实际油水分离过程中有很大应用潜力。     

热定形对并列复合再生聚酯短纤维性能的影响

以再生聚酯瓶片料和泡料混合料为原料进行并列复合纺丝,并经后纺工艺处理得到并列复合再生聚酯短纤维。通过对纤维进行干热定形,研究热定形温度、时间对并列复合再生聚酯短纤的强伸性能、卷曲性能和热收缩性能的影响。结果表明:聚酯短纤维的断裂强度和断裂伸长率随着热定形温度升高而增大;断裂强度随热定形时间的延长逐渐下降,断裂伸长率先增大后减小,在20 min时达到最大值,为17.4%,声速取向因子则随着热定形时间的延长呈现下降趋势。纤维的卷曲性能随着热定形温度的升高而改善,较短的时间内,纤维的卷曲性能已经达到最佳;热定形温度的升高使纤维的热收缩率增大;并列复合再生聚酯短纤维的最佳热定形温度是140～160℃,最佳定形时间为10 min。     

轻量级电驱水下机械臂设计与运动学分析

设计了一款四功能水下机械臂，并对驱动关节进行了模块化设计和密封测试。为了保证水下机械臂的可靠性，在水下机械臂水下测试前，进行了强度校核和模态分析。为了验证水下机械臂的性能，通过 Ｄ- Ｈ法建立了其运动学模型，在此基础上分别进行了正、逆运动学分析。此外还采用蒙特卡洛法，在正运动学分析的基础上研究了工作空间特点。研究结果将为进一步的运动控制提供有益的帮助。     

Enhanced airborne sound absorption effect in poly(vinylidene fluoride)/(K0.5Na0.5)NbO3-nanofiber composite foams

Introducing electrical conductive function to discharge local piezoelectric effect is found effective for improving airborne sound absorption performance. In this work, instead of conductive fillers, a composite with two piezoelectric materials with opposite piezoelectric responses was explored aiming at enhanced sound absorption effect. Open-cell poly(vinylidene fluoride)/(K_0.5Na_0.5)NbO₃ (PVDF/KNN)-nanofiber composite foams were proposed and investigated for airborne sound absorption purpose. Structural and thermal analyses showed that the KNN nanofibers were well dispersed in the PVDF matrix and enhanced the degree of crystallinity of polar phase of PVDF. Significantly enhanced airborne sound absorption over a broad frequency range was observed in the PVDF/KNN-nanofiber composite foams, with increasing KNN nanofibers. One possible mechanism for the improved sound absorption with the piezoelectric KNN nanofibers with positive piezoelectric coefficient added in the PVDF matrix with negative piezoelectric coefficient is that electrical discharge could be facilitated for energy dissipation with the opposite charges generated through the piezoelectric effects in the two phases with opposite polarity. The experimental results show that the open-cell PVDF/KNN-nanofiber composite foams are promising for broadband airborne sound absorption application, and our analysis shed a light on the strategy in designing piezoelectric composite foam with high sound absorption performance.