基于热红外图像处理技术的农作物冠层识别方法研究

为解决农作物冠层热红外图像边缘灰度级分布不均且噪声较大,而传统图像分割方法难以实现其目标区域有效识别的难题,以苗期红小豆冠层热红外图像为研究对象,将模糊神经网络和仿射变换有机结合,提出了基于热红外图像处理技术的农作物冠层识别模型。首先利用五层线性归一化模糊神经网络的自适应特性,选取高斯隶属度函数,自动计算冠层可见光图像识别的推理规则,有效地分割了可见光图像中的冠层区域。通过分析3种分割指标和熵,定量评价可见光图像冠层分割质量。网络迭代38次时,误差精度为0.000 952,该算法平均有效识别率为96.13%,获取可见光冠层图像的像元信息熵值范围为2.454 4~5.198 7,与标准算法所得冠层图像的像元信息熵仅相差0.245 9。然后以取得可见光图像的冠层有效区域为参考图像,采用仿射变换算法,调整优选平移、旋转、缩放等图像变换因子,配准原始热红外图像,提出了基于仿射变换的冠层热红外图像识别方法。对于初始温度范围值在16.35~19.92 ℃的农作物热红外图像,计算选取旋转幅度为1.0和缩放因子为0.9时,作为异源图像的最优配准参数,获取目标图像的最大温差为3.17 ℃,相对于原图像的平均温度值由18.711 ℃下降至17.790 ℃,进而实现了基于热红外图像处理技术的农作物冠层识别。最后以熵的互信息作为监督指标,对农作物冠层热红外图像识别方法进行评价。提出的冠层热红外图像识别方法,所获取的目标图像与初始热红外图像的平均互信息为4.368 7,标准目标图像和初始热红外图像的平均互信息为3.981 8,二者仅相差0.486 9。同时,两种冠层热红外图像的平均温度差值为0.25 ℃,高效消除了原始热红外图像的背景噪声。结果表明本研究方法的有效性和实用性,能够为应用热红外图像反映农作物生理生态信息特征指标参数提供技术借鉴。     

Modélisation du comportement thermique d'un espace urbain: Calcul de la réponse en température des structures et de l'air ambiant

This paper presents theoretical modelling works on the thermal behaviour of an urban space (street). A calculation code, named Codyflow, has been developed in order to simulate the thermal response of an urban system to climatic solicitations. The model allows us to take into consideration, on the one hand, the radiative flux soliciting the urban canyon, on the other hand, heat transfers inside the system. The airflow solicitation, which plays a part in the convective exchanges and in the ambient air temperature, is calculated by the CFD code Fluent. Some simulation results, obtained by the code Codyflow, are presented. They bring to the fore the influence of many factors on the thermal response of the urban canyon: the geometrical configuration, the used materials (through their thermophysical characteristics) and the airflow solicitation. These results allow us to predict the thermal behaviour of urban spaces for the benefit of architects and urban designers in the conception phase of an urban plan.     

Detection of buried objects using infrared imaging of impulse laser heated soils

In this work we examine the time-dependent aspects of thermal imaging of buried metallic and non-metallic objects, when impulse heating of the soil surface is performed. A 1500 W carbon dioxide laser is used to provide the thermal impulse. Time-dependent thermal images are obtained using an infrared focal plane array camera and a gated data acquisition/analysis system. Experimental studies are conducted as a function of object material, size, and burial depth, and incident laser fluence. Quantitative evaluation of the surface temperature differentials as a function of time are presented.     

红外图像伪彩色编码和处理

利用红外热成像系统可将物体的热分布转化为可视图像，并在监视器上以灰度级或伪彩色显示出来，从而得到被测目标的温度分布场。根据热成像测温原理以及红外图像的特点，在对室温热成像系统研究的基础上，对红外图像伪彩色编码进行了研究，提出一种新的伪彩色编码方法，即自动阈值法。利用自动阈值法可以在室温环境下，对目标的温度及其分布进行测量。在TMS320C6202和FPGA室温热成像系统中对提出的方法进行了验证。实验结果表明，该方法可使红外图像层次分明，容易分辨出不同的温度区域。     

Improving spatio-temporal resolution of infrared images to detect thermal activity of defect at the surface of inorganic glass

This paper proposes a noise suppression methodology to improve the spatio-temporal resolution of infrared images. The methodology is divided in two steps. The first one consists in removing the noise from the temporal signal at each pixel. Three basic temporal filters are considered for this purpose: average filter, cost function minimization (FIT) and short time Fast Fourier Transform approach (STFFT). But while this step effectively reduces the temporal signal noise at each pixel, the infrared images may still appear noisy. This is due to a random distribution of a residual offset value of pixels signal. Hence in the second step, the residual offset is identified by considering thermal images for which no mechanical loading is applied. In this case, the temperature variation field is homogeneous and the value of temperature variation at each pixel is theoretically equal to zero. The method is first tested on synthetic images built from infrared computer-generated images combined with experimental noise. The results demonstrate that this approach permits to keep the spatial resolution of infrared images equal to 1 pixel. The methodology is then applied to characterize thermal activity of a defect at the surface of inorganic glass submitted to cyclic mechanical loading. The three basic temporal filters are quantitatively compared and contrasted. Results obtained demonstrate that, contrarily to a basic spatio-temporal approach, the denoising method proposed is suitable to characterize low thermal activity combined to strong spatial gradients induced by cyclic heterogeneous deformations.     

几种直方图变换法在输电接头红外热像增强中的对比研究

由红外热像仪直接得到的热像会受很多外来因素的干扰而致使图像模糊不清、对比度差,由此在故障诊断过程中,需要对热像进行后续增强处理。总结了影响热故障红外检测准确性的外部因素,分析了输电接头热像的基本特征,采用了基于平方根灰度变换直方图修正和全局均衡化两种方法对热像进行了增强处理。在分析上述两种方法优劣的基础上,提出了一种基于线性平滑滤波的局部均衡化方法。结果表明,基于线性平滑滤波的局部均衡化方法对于处理输电接头热像更有效。     

潜在低秩表示框架下基于卷积神经网络结合引导滤波的红外与可见光图像融合

为提高融合图像的可视性,解决传统红外与可见光图像融合算法中存在的边缘特征缺失、细节模糊的问题,提出了一种潜在低秩表示框架下基于卷积神经网络结合引导滤波的红外与可见光图像融合算法。该算法首先利用潜在低秩表示对源图像进行分解,得到源图像的低秩分量和显著分量。其次,利用卷积神经网络根据源图像的特征信息,得到权值图。再次,通过引导滤波算法对权值图进行边缘锐化,然后再将优化后的权值图分别与源图像的低秩分量和显著分量融合,得到融合图像的低秩分量和显著分量。最后,将融合图像的低秩分量和显著分量叠加,得到最终的融合图像。实验结果表明,该算法在主观评价和客观指标上均优于传统的红外与可见光图像融合算法。     

Multi-band infrared imaging radiometer

A novel approach to multi-spectral thermal imaging using staring system has been developed. An original optical assembly has been designed set up and tested that allows to acquire simultaneously four images of the same target in different spectral bands using a single commercial IRFPA sensor and a single commercial focusing lens. The spectral and spatial separation of the four images is completely obtained by means of filters and refractive optics without any kind of scanning device or other moving part. The multi-spectral optic presented in this paper operates in the spectral region of middle infrared (3–5 μm). Spectral filters are used to select the appropriate bands of interest and they are easily interchangeable.     

Warping-based co-registration of thermal infrared images: Study of factors influencing its applicability

A relevant issue for processing biomedical thermal imaging data is the availability of tools for objective and quantitative comparison of images across different conditions or subjects. To this goal, a solution can be offered by projecting the thermal distribution data onto a fictitious template to obtain a common reference for comparison across cases or subjects.In this preliminary study, we tested the feasibility of applying a warping procedure on infrared thermal images. Fifteen thermal images of checkerboard were recorded at three different distances and five different angles in order to evaluate which factor mostly influences the warping accuracy.The accuracy of three different warping transformation models (local weighted mean (LWM), polynomial, affine) was tested by comparing the positioning error between users’ selected fiduciary points on each thermal image and their corresponding reference position assigned on the template image.Fifteen users, divided into three groups upon on their experience in thermal imaging processing, participated in this study in order to evaluate the effect of experience in applying a warping procedure to the analysis of thermal infrared images.The most relevant factor influencing the positioning and thermal errors is the acquisition distance, while the users’ level of experience and the inclination angle do not seem to play the same importance. Comparing the three transformations, the LWM seems to be the best in terms of minimizing the two categories of errors. This preliminary work helps to understand the limits and the possibilities of applying warping techniques for objective, quantitative and automatic thermal image comparisons.     

A Generative Adversarial Network for Infrared and Visible Image Fusion Based on Semantic Segmentation

This paper proposes a new generative adversarial network for infrared and visible image fusion based on semantic segmentation (SSGAN), which can consider not only the low-level features of infrared and visible images, but also the high-level semantic information. Source images can be divided into foregrounds and backgrounds by semantic masks. The generator with a dual-encoder-single-decoder framework is used to extract the feature of foregrounds and backgrounds by different encoder paths. Moreover, the discriminator’s input image is designed based on semantic segmentation, which is obtained by combining the foregrounds of the infrared images with the backgrounds of the visible images. Consequently, the prominence of thermal targets in the infrared images and texture details in the visible images can be preserved in the fused images simultaneously. Qualitative and quantitative experiments on publicly available datasets demonstrate that the proposed approach can significantly outperform the state-of-the-art methods.     

太赫兹激励的红外热波检测技术

本文的目的在于探索一种新的适用于红外热波检测技术的热激励方式——太赫兹(THz)热激励. 文中介绍了THz波周期性热激励的热传导理论模型; 尝试利用返波振荡器(返波管backward wave oscillator, BWO)太赫兹源对一块碳纤维基底吸波涂层板进行周期性THz热激励, 红外热像仪连续观测和记录试件表面温场变化, Canny边缘算法处理热图像显示缺陷; 检测结果与闪光灯脉冲激励的结果进行比较, 讨论了太赫兹波激励红外热波检测技术可能的优势. 实现了THz技术与红外热波无损检测技术的结合.     

利用QuickBird数据进行热红外遥感成像模拟的研究

针对基于机载热像仪图像或随机产生仿真场景的早期热红外成像模拟系统,提出了对高分辨率QuickBird图像进行星载热红外成像模拟的方法。基于高分辨率QuickBird图像,采用ISODATA法进行非监督分类,得到土地分类专题图,通过地物光谱数据库获得场景地物温度及发射率,从而可获得地面辐亮度图像;利用大气辐射传输软件MODTRAN对其进行大气修正,结合卫星遥感器辐射定标系数,最终可得到卫星遥感器输出图像。仿真结果表明,利用QuickBird进行热红外遥感成像模拟是一种全新而有益的尝试,对伪装效果检验等具有一定的参考价值。     

Advanced Driving Assistance Based on the Fusion of Infrared and Visible Images

Obtaining key and rich visual information under sophisticated road conditions is one of the key requirements for advanced driving assistance. In this paper, a newfangled end-to-end model is proposed for advanced driving assistance based on the fusion of infrared and visible images, termed as FusionADA. In our model, we are committed to extracting and fusing the optimal texture details and salient thermal targets from the source images. To achieve this goal, our model constitutes an adversarial framework between the generator and the discriminator. Specifically, the generator aims to generate a fused image with basic intensity information together with the optimal texture details from source images, while the discriminator aims to force the fused image to restore the salient thermal targets from the source infrared image. In addition, our FusionADA is a fully end-to-end model, solving the issues of manually designing complicated activity level measurements and fusion rules existing in traditional methods. Qualitative and quantitative experiments on publicly available datasets RoadScene and TNO demonstrate the superiority of our FusionADA over the state-of-the-art approaches.     

基于无人机红外光谱技术的影像数据匹配方法研究

无人机加载红外光谱载荷对区域内影像进行获取现已成为遥感领域一种重要的技术手段,可通过对携带位置信息的影像进行分类提取,得到植被盖度、温度指数等一系列因子指标。利用FREE BIRD(自由鸟)小型低空无人机系统挂载Tetracam红外相机(310万像素)对新疆玛纳斯县一河道进行影像获取。无人机飞行面积约为20.5 km2,为了得到更加精确的植被、温度等因子,需要对无人机红外影像进行配准,通过优化SIFT匹配参数和RANSAC粗差剔除后,获取了可靠的匹配结果,即经过算法匹配后的影像与原影像进行了误差比对,能够满足后期的应用需要, 这也是本文的创新点之一。将影像进行配准后进行二维影像拼接,将多张红外影像按照航向重叠度不低于60%,旁向重叠度不低于50%的概率进行拼接,得到拼接后的红外影像图。另外比较了SIFT和SUFT两种算法,利用优化的SIFT算法及改进的FLIR传感器获取1 600张热红外影像,利用地面同步测量数据对拼接后的红外影像进行算法匹配,并利用ENVI(完整的遥感图像处理平台)软件进行温度及植被盖度的影像反演,得到了研究区域的单一影像及红外影像的温度反演图及植被反演图。通过对两种算法的对比得到更加优化的算法模型,并对该模型进行回归分析和精度检验,得到该模型的相关系数R2为0.767,匹配精度为81.51%,模型精度较高。本模型的建立对日后无人机红外影像的配准及提取反演奠定了理论和实践基础。     

Thermographic diagnosis system and imaging algorithm by distributed thermal data using single infrared sensor

An infrared diagnosis device provides two-dimensional images and patient-oriented results that can be easily understood by the inspection target by using infrared cameras. However, this device has disadvantages such as large size, high price, and inconvenient maintenance. In this regard, this study has proposed a small diagnosis device for body heat using a single infrared sensor and implementing an infrared detection system using a single infrared sensor and an algorithm that represents thermography using the obtained data on the temperature of the point source. The developed system had a temperature resolution of 0.1 °C and reproducibility of ±0.1 °C. The accuracy was 90.39% at the error bound of ±0 °C and 99.98% at that of ±0.1 °C. To evaluate the proposed algorithm and system, the infrared images of the camera method were compared. To verify the device’s clinical applicability, thermal images with clinical meaning were obtained from a patient who had lesions.     

基于EPICS的颗粒传热实验测控系统应用研究

颗粒传热实验测控系统是为颗粒换热平台设计的测控系统。该系统基于分布式系统EPICS架构设计,通过对NI硬件平台、红外热像仪设备以及Oracle数据库进行集成,实现了颗粒换热平台实验的在线采集、监测、控制与实时存储。由于红外热像仪具有灵敏度高、测温范围广、非接触测温等优点,因此在该测控系统中用于测量颗粒在换热器出口处的温度。本文使用SDK将其集成到EPICS系统中,利用红外图像更加直观地呈现出颗粒的实时温度分布,并提取红外图像中的数据,与热电偶的测量数据对比,进一步检测利用红外热像仪作为一种颗粒实时温度测量设备的可行性。实验结果表明,集成后的红外热像仪操作软件能够实时提取并存储红外图像中的温度数值,并且采用EPICS架构设计、开发的颗粒传热实验测控系统完全可以满足颗粒换热平台的数据采集与存储需求。     

自适应参考图像的可见光与热红外彩色图像融合算法

可见光与红外热图像的彩色图像融合技术是现今国内外高性能夜视技术发展的重要方向之一,该技术有效提高了人们对目标的探测和场景理解能力。目前常用的色彩传递算法多属于基于单幅参考图像的全局色彩传递算法,彩色融合图像的色调受到参考图像的影响较大,在实际应用中难以保证对各类场景的适应性。针对常规YUV空间色彩传递彩色图像融合算法的环境适应性问题,通过对植物、城镇和海天三类典型场景的分类与统计,发现了典型场景在UV通道的均值和标准差具有的较为明显的分类特性,由此提出了一种基于UV通道均值和标准差的自适应参考图像构造方法,使得可见光与热红外彩色图像融合算法具有较常规算法更好的环境适应性,融合图像的色彩具有较好的自然感,且算法处理量较小,对现有实时硬件融合处理算法的运算速度影响不大,是一种环境适应性强的自然感彩色融合处理算法。     

Thermal heat radiation,near-field energy density and near-field radiative heat transfer of coated materials

We investigate the thermal radiation and thermal near-field energy density of a metal-coated semi-infinite body for different substrates. We show that the surface polariton coupling within the metal coating leads to an enhancement of the TM-mode part of the thermal near-field energy density when a polar substrate is used. In this case the result obtained for a free standing metal film is retrieved. In contrast, in the case of a metal substrate there is no enhancement in the TM-mode part, as can also be explained within the framework of surface plasmon coupling within the coating. Finally, we discuss the influence of the enhanced thermal energy density on the near-field radiative heat transfer between a simple semi-infinite and a coated semi-infinite body for different material combinations.     

Nanometer-scale probing of optical and thermal near-fields with an apertureless NSOM

We have used a home-made apertureless near-field scanning optical microscope (ANSOM) for mapping nanometric steps between SiC and gold regions under visible (λ=655 nm) and infrared (λ=10.6 μm) illumination. The images, obtained with a signal demodulation at the tip oscillation frequency and at higher harmonics, clearly show optical contrasts with a subwavelength resolution of about 30 nm. Other images recorded in the visible on a YBa₂Cu₃O₇ crystal indicate that the tip used in our experiments is able to reveal polarization effects. We also present a near-field thermal optical microscope (NTOM) which operates without any external illumination. In this new kind of microscope, the laser source which is usually used to excite the evanescent waves, is replaced by a simple heating of the sample. The electromagnetic radiation locally scattered by the tip comes from the thermal radiation. Our results with this new technique prove a 200 nm lateral resolution.     

Active infrared thermographic inspection technique for elevated concrete structures using remote heating system

Conventionally, the inspection of elevated concrete structures requires the use of scaffolding or an aerial truck. In this study, elevated railway structures constructed of reinforced concrete were inspected using active infrared thermography. The inspection area corresponded to half of the middle slab covering an area of 16.8 m²; one inspection was carried out that took about 15 min. A remote heating system consisting of a 6-kW air-cooled xenon arc lamp and a scanner system was developed to detect hidden defects in elevated concrete structures without the need for an aerial truck or scaffolding. The generation of a thermal image and irradiation are carried out simultaneously by the beam scanning. High-contrast infrared thermal images can be obtained by the simple image processing procedure that is proposed.