融合深度学习与成像模型的水下图像增强算法

水下机器人的视觉感知功能因受到水下环境因素的影响,面临着图像质量降低的挑战,如图像颜色畸变、整体色调偏绿、偏蓝、对比度较低、细节较为模糊等。提出一种结合深度学习方法与物理成像模型的新型水下图像增强算法,通过构建包含扩张卷积和带参数激活函数的神经网络,进行背景散射光和直接传输映射的估计,并结合成像模型的数学表达进行重建运算得到增强后图像。实验结果表明,与UDCP、IBLA、GLNet等典型图像增强算法相比,该算法具有更快的运算速度,且能够消除水下环境因素带来的影响,丰富图像色彩的同时能增强各类细节,在峰值信噪比指标和结构相似度指标上取得了较大值。此外,增强后的图像在特征点匹配实验中获得了更好的匹配效果。     

Enhancement of Upconverted Fluorescence by Interference Layers

Upconverting nanoparticles show potential applications in the field of photovoltaics and array‐based detection devices. While fluorescence enhancement using interference of incident radiation is well known in Stokes‐shift type systems such as fluorescent dyes; the effect of such interference geometry in nonlinear Anti‐Stokes type emission, such as in upconversion rare earth photophysics is demonstrated for the first time. This work describes in detail the influence of the interference modulation on both the excitation (interion energy transfer) and radiative decay with nonradiative decay processes active between emissive levels. These effects are illustrated in the thickness dependence of the decay rate and rise time. Single particle upconverted spectra and time‐resolved measurements show concurrent optimization of the infrared absorption and emission at 540 and 650 nm, with an average enhanced emission of 20 times at λ = 540 and 45 times at λ = 650 nm, dependent on the interference layer thickness and on the excitation intensity. The experimental results are correlated with finite element modeling. Both experiments and calculations show emission enhancement at an interference layer thickness of about 740 ± 20 nm, where such tolerance and the planar design, leads to ease in implementation in applications.     

An adaptive method based on the improved LPA-ICI algorithm for MRI enhancement

Various diseases are diagnosed using medical imaging used for analysing internal anatomical structures. However, medical images are susceptible to noise introduced in both acquisition and transmission processes. We propose an adaptive data-driven image denoising algorithm based on an improvement of the intersection of confidence intervals (ICI), called relative ICI (RICI) algorithm. The 2D mask of the adaptive size and shape is calculated for each image pixel independently, and utilized in the design of the 2D local polynomial approximation (LPA) filters. Denoising performances, in terms of the PSNR, are compared to the original ICI-based method, as well as to the fixed sized filtering. The proposed adaptive RICI-based denoising outperformed the original ICI-based method by up to 1.32 dB, and the fixed size filtering by up to 6.48 dB. Furthermore, since the denoising of each image pixel is done locally and independently, the method is easy to parallelize.     

Crystallization‐Induced Emission Enhancement of a Deep‐Blue Luminescence Material with Tunable Mechano‐ and Thermochromism

Organic luminescent materials with the ability to reversibly switch the luminescence when subjected to external stimuli have attracted considerable interest in recent years. However, the examples of luminescent materials that exhibit multiresponsive properties are rarely reported. In this work, a new stimuli‐responsive dye P1 is designed and synthesized with two identical chromophores of naphthalimide, one at each side of an amidoamine‐based spacer. This amide‐rich molecule offers many possibilities for forming intra‐ and intermolecular hydrogen bond interactions. Particularly, P1 has an intrinsic property of cocrystallizing with methanol. Compared with the pristine P1 sample, the as‐prepared two‐component cocrystalline material displays an exceptive deep‐blue emission, which is extremely rare among naphthalimide‐based molecules in the solid state. Furthermore, the target material exhibits an obvious mechanochromic fluorescent behavior and a large spectral shift under force stimuli. On the other hand, the cocrystalline material shows an unusual “turn off” thermochromic luminescence accompanied by solvent evaporation. Moreover, using external stimuli to reversibly manipulate fluorescent quantum yields is rarely reported to date. The results demonstrate the feasibility of a new design strategy for solid‐state luminescence switching materials: the incorporation of solvents into organic compounds by cocrystallization to obtain a crystalline state luminescence system.     

Reproducible Enhancement of Fluorescence by Bimetal Mediated Surface Plasmon Coupled Emission for Highly Sensitive Quantitative Diagnosis of Double‐Stranded DNA

Plasmonic enhancement of fluorescence from SYBR Green I conjugated with a double‐stranded DNA (dsDNA) amplicon is demonstrated on polymerase chain reaction (PCR) products. Theoretical computation leads to use of the bimetallic (Au 2 nm–Ag 50 nm) surface plasmons due to larger local fields (higher quality factors) than monometallic (Ag or Au) ones at both dye excitation and emission wavelengths simultaneously, optimizing fluorescence enhancement with surface plasmon coupled emission (SPCE). Two kinds of reverse Kretschmann configurations are used, which favor, in signal‐to‐noise ratio, a fluorescence assay that uses optically dense buffer such as blood plasma. The fluorescence enhancement (12.9 fold at maximum) with remarkably high reproducibility (coefficient of variation (CV) < 1%) is experimentally demonstrated. This facilitates credible quantitation of enhanced fluorescence, however unlikely to obtain by localized surface plasmons. The plasmon‐induced optical gain of 46 dB due to SPCE‐active dye molecules is also estimated. The fluorescence enhancement technologies with PCR enables LOD of the dsDNA template concentration of ≈400 fg µL^?1 (CV < 1%), the lowest ever reported in DNA fluorescence assay to date. SPCE also reduces photobleaching significantly. These technologies can be extended for a highly reproducible and sufficiently sensitive fluorescence assay with small volumes of analytes in multiplexed diagnostics.     

基于YOLOV4的工件表面质量在线检测方法研究

目的 提升自动化产线上工件表面微小缺陷的检测精度和检测速度。方法 首先,在预处理阶段提出采用CutMix的数据增强方法,增加训练样本的多样性,提高模型的鲁棒性和泛化能力,避免训练模型产生过拟合;使用K–means++聚类算法生成边界候选框,以适应不同尺寸的缺陷,并较早地筛选出更精细的特征。其次,借助CSP Darknet53网络及SPP模块提取输入原始图像的特征,通过训练获得针对工件表面质量的在线检测模型,提升YOLOV4缺陷位置检测及识别的精度。结果 实验结果表明,文中所提出的基于YOLOV4的工件表面质量在线监测方法的预测精度达到97.5%,检测速度达到32.8帧/s,均优于同类的深度学习算法。以贵州某航空工业产品的自动化产线作为实验平台验证了所提方法的可行性和有效性。结论 该方法具备结构简单清晰、自适应性强等优点,检测精度和速度均满足工业场景需求,可以将其用于产品表面质量的在线检测。     

Medical image contrast enhancement based on improved sparrow search algorithm

The collection or transmission of medical images is often disturbed by various factors, such as insufficient brightness and noise pollution, which will result in the deterioration of image quality and significantly affect the clinical diagnosis. To improve the quality of medical images, a contrast enhancement method based on improved sparrow search algorithm is proposed in this paper. The method is divided into two steps to enhance the medical images. First, a new transform function is introduced to improve the brightness or contrast of medical images, and two parameters in the transform function are optimized by the improved sparrow search algorithm. Second, adaptive histogram equalization method with contrast limited is used to equalize the result image of the previous step to make the pixel distribution of the image more uniform. Finally, a large number of experiments and qualitative and quantitative analyses were conducted on the common data sets. The analysis results demonstrate that the presented approach outperforms some existing medical image processing approaches.     

Parallel algorithms for image enhancement and segmentation by region growing, with an experimental study

This paper presents efficient and portable implementations of a powerful image enhancement process, the Symmetric Neighborhood Filter (SNF), and an image segmentation technique that makes use of the SNF and a variant of the conventional connected components algorithm which we call -Connected Components. We use efficient techniques for distributing and coalescing data as well as efficient combinations of task and data parallelism. The image segmentation algorithm makes use of an efficient connected components algorithm based on a novel approach for parallel merging. The algorithms have been coded in Split-C and run on a variety of platforms, including the Thinking Machines CM-5, IBM SP-1 and SP-2, Cray Research T3D, Meiko Scientific CS-2, Intel Paragon, and workstation clusters. Our experimental results are consistent with the theoretical analysis (and provide the best known execution times for segmentation, even when compared with machine-specific implementations). Our test data include difficult images from the Landsat Thematic Mapper (TM) satellite data.Also affiliated with the Department of Electrical Engineering.Also affiliated with the Department of Computer Science and the Center for Automation Research.     

基于二阶梯度的图像增强算法研究

针对PCB图像的特点对图像进行增强处理,使得图像的边缘更加突出,灰度的对比度增加,同时滤除靠近边缘的噪声,为后续图像分割时更有效的选取阈值打下良好基础。使用二阶梯度算子获得图像的增强效果,实验表明,不仅解决了边缘附近噪声去除问题,而且有效地突出了边缘信息,图像的黑白对比分明,尤其是目标图像靠近边缘部分灰度值明显高于其他的非边缘区域。