基于有理函数的遥感水体污染图像的自适应插值

针对遥感图像的分辨率和显示质量时常不能同时达到需求的问题,提出一种基于有理函数模型的自适应插值算法(adaptive rational interpolation method,ARIM)。该模型不但具有简洁的表达式而且可以在不改变插值数据的前提下通过参数来调整插值曲面的弯曲度。算法充分考虑到了遥感水体污染图像边缘区域的方向信息,让处理过的遥感图像更加符合人眼视觉系统的特性。通过等值线将图像自适应地分解为平滑区域和非平滑区域,在不同的区域分别用不同的算法进行插值,提升了插值速度。另外,模型参数的最优化也有效提高了插值精度。实验结果表明,该算法不但提升了遥感图像的分辨率,而且在主观(目视解释)和客观(峰值信噪比)质量上具有一定的优势。

农田覆膜播种中白色污染图像盲复原方法研究

图像盲复原能够恢复图像里有价值的信息并改善图像质量,因此该课题在图像处理领域具有较高的研究性和实用价值。在农田覆膜播种白色污染图像恢复中,该技术的应用可获取高质量图像,为白色污染治理提供依据,具有一定现实意义。首先,因为稀疏特性表现在白色污染自然图像边缘,可以利用一种权重的全变差范数对图像实现正则化约束,考虑到运动模糊点扩散函数的特点,把运动模糊函数的连续平滑性及稀疏性运用到图像正则化约束分析中来,达到白色污染自然图像边缘稀疏和锐化增强的效果,这样对图像复原更为有利;其次,以图像正则化约束分析为基础,依据图像先验信息和正则化特点,利用改进版Bregman迭代法建立图像盲复原代数函数,为简化求解在函数中加入惩罚和分裂因子,可以有效获取高质量的白色污染盲复原图像。仿真实验证明,运用本文所述方法能获得更高质量的农田覆膜播种中白色污染图像,为治理污染提供信息资料。

基于多级中值滤波的严重椒盐噪声污染图像恢复

文章提出了一种用于基于多级中值滤波的严重椒盐噪声污染图像恢复的方法。该算法首先用多级中值滤波对图像进行滤波,然后用噪声定位技术将滤波后的图像像素点分为噪声和信号点两类,建立噪声标记矩阵,进而对噪声点进行噪声消除,对非噪声点保持原值不变,从而实现噪声污染图像恢复。实验结果表明,本算法对于噪声浓度大于40%的污染图像去除噪声及保留细节等方面的效果非常显著,尤其对于严重椒盐噪声污染图像。

融合显著信息的白色污染图像自动标注算法

针对传统方法无法精准获取目标信息,导致标注结果不精准的问题,提出融合显著信息的白色污染图像自动标注算法。运用RC算法提取显著性信息,融合物体级别先验信息,增强显著性的目标一致性。利用全变差范数对图像实施正则化约束,将运动模糊点扩散函数具备的稀疏性和连续平滑性应用在图像正则化约束分析中,获取图像底层特征信息,再根据FCM算法,得到各个模糊组的聚类中心,通过遗传算子消除聚类中心收敛至极值的现象,反复搜索获得最优解。根据SimMSVM建立多类支持向量机最后引入损失函数实施优化,获得SimMSVM的类判别函数,经过逐步运算,最终实现白色污染图像的自动标注。实验结果表明:所提算法具有较高适用性、有效性以及准确性。

同态滤波和中值滤波级联的噪声污染激光图像恢复

面对激光图像中噪声污染严重,去噪不彻底的问题,提出同态滤波和中值滤波级联的噪声污染激光图像恢复方法。首先建立矢量窗口检测激光图像中存在的噪声,根据像素点灰度平均值将检测到的噪声分为离散噪声和聚集噪声。然后采用同态滤波降噪方法消除激光图像中的离散噪声,同时利用改进中值滤波方法消除激光图像中的聚集噪声,最后将两种降噪方法相结合对噪声污染激光图像实现恢复工作。实验结果证明,所提方法散斑噪声指数较小,边缘保持指数较大,图像恢复时间较快,去噪效果明显。

光学相干层析成像技术在膜污染研究中的应用

简述了光学相干层析成像技术(Optical coherence tomography,OCT)的成像原理和监测优势;由于可快速生成高分辨率的污染层二维和三维图像,可用于表征污染层形成过程､定量分析污染层､表征膜内部污染与润湿､耦合深度学习模型预测膜污染,促进对膜污染形成过程和机制的理解;总结了OCT在膜污染分析中的主要问题,并对未来的研究进行了展望。

油液污染颗粒边缘检测算法研究

边缘检测是油液污染颗粒图像处理的重要组成部分,其性能直接影响油液污染度光电检测系统的精度和性能。对Sobel、LoG、Canny和最佳阈值迭代算法进行了分析,对油液污染颗粒边缘检测进行了实验研究。结果表明,最佳阈值迭代算法在油液污染颗粒图像中的检测性能较好。

数字媒体图像多特征反差增强方法仿真

由于图像在生成的过程中会有相对的噪声产生,而噪声会直接影响该图像中的多特征信息,导致信息表达不完全等情况。基于此提出了数字媒体图像多特征反差增强方法。首先将需要对受噪声污染的图像进行去噪;其次运用偏微分方程(PDE)方法,构建出一种PDE数学模型-TV模型,在对图像去噪的同时,保持原图像的边缘轮廓信息,并且运用Split Bregman算法,并引入辅助变量转换出分段线性函数的泛函极值;在偏微分方程的基础上,运用改进后的反差增强方法,根据直方图的均衡化结果利用累积分布函数的方法,来当做图像灰度的变换函数,以达到合理的调整固定动态范围,且让图像整体视觉效果增强、提高图像中多特征信息的目的。通过仿真结果可以得知,提出的方法可以处理图像中噪声的影响,并且可以将图像中的多特征信息进行有效的反差增强处理,且具有直方图均衡化结果优秀、图像特征增强效率的优点。

基于图像滤波预处理的卷积神经网络汉字识别

利用卷积神经网络可以对破损、模糊不清的文字进行有效识别。为了实现速度快、精度高等优点,通过卷积神经网络中的LeNet-5网络模型对手写汉字图像进行识别。首先,在模拟写字板中建立手写汉字的图像数据集,搭建并训练卷积神经网络模型保存图像特征;然后对输入的手写汉字图像进行模拟污染并采用7种滤波去噪方式;最后对加噪、滤波处理后的图像进行识别,对比不同滤波处理的准确性。实验结果可表明,该方法能高效、稳定地从有噪声图像中识别出文字,同时经高斯滤波与PCA滤波处理后的图像识别精确度更高。

基于深度学习的光伏发电技术研究

为了提高光伏发电系统运行效率,以阐述光伏发电技术概述为基础,分析光伏发电技术运行原理,明确其主要特征,为研究人员收集光伏发电系统运行数据打下坚实的基础。同时,进行光伏阵列污染图像分类模型仿真分析,通过建立光伏阵列污染图像分类模型,科学预测未来的天气条件,并根据预测结果调整光伏发电系统的工作状态,以提高其运行效率。实验结果表明:光伏阵列污染图像分类模型具有较高的准确性和可靠性;与传统方法相比,该模型在发电量预测方面具有良好的表现,能适应不同的天气条件和环境变化。未来,将进一步完善该模型,并探索其他深度学习算法的应用,进一步提高光伏发电系统的运行效率。

集成光伏发电功率预测技术研究

太阳能光伏发电输出功率在气象要素变化下多表现出不稳定情况。故研究建立了门控循环单元与光梯度增强机相结合的集成技术功率预测模型。仿真结果表明,该集成模型的平均时间成本小于200s,且其对不同天气状况下的发电功率预测精度较高。本文提出的集成光伏发电功率预测技术能为光伏发电系统稳定性、效益性、安全性的提升贡献一份力量。

Image-based Water Level Measurement Method under Stained Ruler

This paper propoes the water level measuring method based on the image, while the ruler used to indicate the water level is stained. The contamination of the ruler weakens or eliminates many features which are required for the image processing. However, the feature of the color difference between the ruler and the water surface are firmer on the environmental change compare to the other features. As the color differeaces are embossed, only the region of the ruler is limited to eliminate the noise, and the average image is produced by using several continuous frames. A histogram is then produced on the height axis of the produced intensity average image. Local peaks and local valleys are detected, and the section between the peak and valley which have the greatest change is looked for. The valley point at this very moment is used to detect the water level. The detected water level is then converted to the actual water level by using the mapping table. The proposed method is compared to the ultrasonic based method to evaluate its accuracy and efficiency on the various contaminated environments.

Repair wind field in oil contaminated areas with SAR images

In this paper,we compared the normalized radar cross section in the cases of oil spill,biogenic slicks,and clean sea areas with image samples made from 11-pixel NRCS average,and determined their thresholds of the NRCS of the synthetic aperture radar. The results show that the thresholds of oil and biogenic slicks exhibit good consistency with the corresponding synthetic aperture radar images. In addition,we used the normalized radar cross section of clean water from adjacent patches of oil or biogenic slicks areas to replace that of oil or biogenic slicks areas,and retrieve wind field by CMOD5.n and compare wind velocity mending of oil and biogenic slicks areas with Weather Research and Forecasting modeled data,from which the root mean squares of wind speed(wind direction) inversion are 0.89 m/s(20.26°) and 0.88 m/s(7.07°),respectively. Therefore,after the occurrence of oil spill or biogenic slicks,the real wind field could be repaired using the method we introduced in this paper. We believe that this method could improve the accuracy in assessment of a real wind field on medium and small scales at sea,and enhance effectively the monitoring works on similar oil or biogenic slicks incidents at sea surface.

Land Parcel Land Use History as a Key to Site Selectionfor Documenting Soil Contamination Risk: a Case Study from Australian Suburbia

In that orcharding in early to mid twentieth century southeastern Australia involved use of certain heavy metal and As compounds in regular pest control spray procedures, some interest attaches to the possibility that these landparcels are underlain by soils with above background Cu, Pb and As levels. Interpretation of Land cover changes allowed land parcels previously occupied by orchards to be identified in the 1950s through time series air photos. A comparison of soil analysis results referring to soil samples from control sites, and from land parcels formerly occupied by orchardists, shows that contamination (above background) levels of cations in the pesticides can be found in the top 6 cm of former orchard soils. It is clear that digital spatial data handling and culturally informed air photo interpretation has a place in soil contamination studies, land use planning (with particular reference to re development) and in administration of public health.

Iterative Adaptive Median Filter for Impulse Noise Cancellation

Based on the characteristics of impulse noises, the authors establish a new filter, Iterative Adaptive Median Filter （IAMF）. Acccording to the characteristics of images polluted by impulse noises, they establish weight function combined with iterative algorithm to eliminate noises. In IAMF filter process, because the noise sixes do not participate in the computation, they do not influence the normal points in the image, therefore IAMF can retain the detail well, maintain the good clarity after processing image, and simultaneously reduce the computation. Experiments showed that IAMF have ideal denoising effect for the images polluted by the impulse noises; especially when the noise rates are more than 0.5, IAMF is mote prominent, even when the noise rotes are more than 0.9, IAMF can achieve a satisfactory results.

Rapideye Images in the Study of Contaminated Areas： The Case of Niccioleta （Tuscansy-ltaly）

Mining activity in Italy has been one of the main productive activities for millennia, particularly in the Tuscany region which has a great mining tradition, unfortunately characterized in the past by a management little interest to environmental problems. The area under study is the disused mine Niccioleta, in Val d＇Aspra, located about 6 km NE of Massa Marittima in the province of Grosseto. The area is characterized by the presence of four major landfills, in which prevail quantitatively fine-grained materials resulting from the treatment by flotation of pyrite. The study of satellite images offers a new approach to the study of environmental problems. The results obtained from the RapidEye images showed the presence of pyrite and chalcopyrite followed from arsenopyrite, as confirmed by the analysis of diffractometer of the samples and by bibliographic data. RapidEye images lend themselves very to be used to monitor areas of disused mining deposits of ores with primary mineralization predominantly sulphides and subject to oxidized characterized by processes of oxidation/dissolution of pyrite sulphide most common and abundant. In fact, the results of this study have highlighted the potential of remote sensing applied to the study of mining areas, noting the possible benefits, both time and cost, which could be obtained by using these techniques.

A triple-exposure color PIV technique for pressure reconstruction

The study developed a triple-exposure color particle image velocimetry(TE-CPIV) technique associated with pressure reconstruction, and validated its feasibility. A light source with the three primary colors of red, green, and blue(R, G, and B) is produced in a time sequence by a liquid crystal display(LCD) projector. Particle images at three different instants under the color illuminations are captured in one snapshot using a color digital single-lens reflex(SLR) camera with a complementary metal-oxide semiconductor(CMOS) sensor. A contamination correction algorithm based on a specific calibration is performed on the different color layers(R layer, G layer, and B layer) of the raw color image to reduce the contaminated intensity of each color illumination on the other two color layers. The corrected intensity generates three new color layers, from which a standard cross-correlation process in the classical PIV method is used to obtain two velocity fields. Eventually, an instantaneous pressure field is reconstructed from the two velocity fields. The feasibility of TE-CPIV was tested by two experiments with a solid body rotation flow and a cylinder wake flow. The results show acceptable accuracy and robustness of the new technique. The idea of the TE-CPIV is believed to provide a simple and effective way of estimating a pressure field with low cost and high convenience.