一种红外成像系统现场性能预测方法

分析了红外成像系统现场性能预测的传统方法,针对传统方法的不足,提出了一种红外成像系统现场性能预测的新方法。该方法从外场实拍热图像中获取目标和背景的平均温差,然后利用TOD准则和目标概率传递函数,得到目标和背景表观温差与红外成像系统探测识别性能的关系,实现典型目标在外场典型环境下的探测识别概率预测。结果表明,该方法与试验数据有一致性。     

基于TOD的红外成像系统作用距离研究

以三角形方位鉴别(TOD)为基础,建立了新的目标获取模型.根据TOD理论,首次结合能量的观点,推导了目标辐射能量通过红外成像系统到达人眼/大脑时可被感知的信噪比.以视觉可感知的阈值信噪比作为理论基础,建立了新的TOD表达式.将目标按照面积元不同的温度空间分布划分为N个等效等边三角形,以辐射原理为基础,建立了人眼观察到目标时TOD与红外成像系统获取目标距离之间的关系式.根据TOD准则,在实例中得出了不同鉴别等级(探测、识别、辨认)时红外成像系统的作用距离.     

三角形方向鉴别阈值性能理论模型研究

阐述了红外成像传感器性能表征的TODT度量方法;依据频谱理论,首先确定了三角形标准靶的空间频谱描述函数;利用匹配滤波器的概念分别推导了人眼/大脑系统通过红外成像传感器可感知的非周期三角形标准靶信号和噪声;以图像视觉感知信噪比为理论基础,首次建立了红外成像传感器的TODT性能模型,并给出了计算机仿真结果.与实验测量数据比较显示了此模型预测红外成像传感器TODT度量曲线的合理性.     

红外目标识别图像复杂度度量方法综述

红外图像复杂度度量方法不仅可以用于描述目标识别面临的复杂场景变化,而且在红外成像系统性能预测与评估、目标识别算法性能对比、建立和改进目标获取性能模型等方面也有广泛而重要的应用。给定了红外目标识别图像复杂度的定义,对该领域近年来最新出现的和部分经典的度量方法进行系统的归纳总结和对比分析,提出了度量方法选择的依据,指出了现有度量方法的缺点和不足,并指出红外目标识别图像复杂度度量未来将向着融合多种特征或者综合多种度量方法的趋势发展。     

机载雷达和红外数据融合的智能目标识别

为了提高机载传感器目标识别系统的性能,提出了利用机载雷达和红外成像传感器数据融合的智能目标识别算法.对红外成像传感器,采用了基于小波矩特征和BP神经网络的目标识别算法,首先提取目标图像的小波矩特征并进行特征选择,然后通过BP神经网络对目标图像进行识别;对雷达传感器,提出了利用模糊推理的目标识别方法,首先选取适当的雷达特征,然后通过模糊推理进行识别:从雷达和红外传感器识别算法分别得到待识别目标所属类别的基本概率分配函数,用D-S证据组合规则将两个基本概率分配函数组合,最终实现了机载雷达和红外传感器的数据融合.仿真结果表明:融合后的识别效果优于单个雷达或红外传感器的识别效果.     

成像目标的红外隐身效果评估

装甲车辆红外隐身效果评估是研制新型装甲车辆,提高其生存力和战斗力的关键技术之一.建立了利用识别概率进行装甲车辆隐身效果评估的方法,该方法通过基于目标与背景纹理相似度的红外图像目标识别概率计算模型,计算模拟红外图像中装甲车辆的识别概率.结果表明该评估方法可以对战车的红外隐身效果进行评估,结合目标与背景红外辐射特性模拟软件可以对研制中的新型装甲车辆的红外隐身方案以及现役型号装甲车辆红外隐身改进措施的隐身效果进行评估.     

红外多谱段小目标识别方法

随着红外诱饵技术的不断发展,对传统的小目标识别提出了新的挑战.通过对飞行目标和红外诱饵光谱特征的分析,利用红外多谱段图像对天空背景中具有伴飞诱饵干扰的小目标进行识别.提出了自适应阈值分割和能量相关滤波相结合的方法对图像进行预处理,再利用光谱角算法对疑似小目标进行识别,增加了目标识别的可靠性,降低了算法的复杂度.与传统的单波段目标识别跟踪算法不同,该算法从光谱特征差异性角度对目标进行识别,识别概率高,运算量小,为红外小目标识别提供了新的解决办法.     

最大互信息在模版匹配中的应用

自动目标识别是进行精确打击的关键技术,其主要实现方法是图像匹配.为提高自动目标识别中图像匹配的精度和速度,提出一种基于最大互信息的快速图像匹配算法.根据定义的图像匹配相似度量,利用灰度信息进行模板匹配,并应用分离图像信息贫乏区的方法提高了算法的图像匹配性能.仿真结果表明,所提方法可有效提高红外图像匹配的实时性和正确性.     

基于热释电红外传感器的运动目标识别方法

为精准地识别运动目标,研究基于热释电红外传感器的运动目标识别方法。构建热释电红外传感器模型,采集红外运动目标图像,通过菲涅尔透镜阵列分割和标记编码空间视觉范围,确定相应运动目标所在地点;将图像进行灰度化、阈值化、小波变换结合自适应增强算法增强处理,利用小波分析法提取图像特征,输入卷积神经网络,输出结果即运动目标识别结果。实验结果表明:所研究方法能够有效识别出运动目标,目标的检测概率近似100%,迭代训练的损失值较低,维持在0.2～0.4区间内,精度值较高,保持在0.8～1.0之间,提取的图像特征值最高可达到1,可实现高精度的目标识别。     

地面红外目标图像识别方法研究

红外成像自动目标识别是精确制导武器的重要研究内容.针对地面复杂背景条件下,目标难以检测识别的问题,在分析地面目标红外图像场景的基础上,首先利用形态学技术对图像进行去噪和抑制背景,采用Otsu法对图像进行阈值分割并用形态学技术消除虚警目标点,用扩展像素标记法对图像进行区域标记,然后提取目标形状、矩和统计分布三类特征作为识别特征,并采用模糊综合评判方法识别目标.对实地拍摄的100幅坦克目标红外图像进行了识别实验,结果表明,该方法在坦克的典型作战环境下具有较强的抗噪性能和抑制背景结构干扰的能力,能有效识别出目标,并且易于用并行处理和硬件实现.     

Field Performance Prediction Method Based On Contrast Metric Of Target Sub Regions

The field performance prediction of infrared imaging system is an important subject in infrared imaging field. In this paper, a novel method for predicting TA (target acquisition) performance is proposed, which combines the TOD (triangle orientation discrimination) threshold curve with an improved contrast metric based on the temperature in target sub regions. Specifically, the target is first subdivided into some sub regions with different target/background differential temperatures according to its thermal structure distribution. The number of equivalent triangle patterns across each sub region is calculated utilizing its effective size, the apparent temperature difference line, and the TOD curve. Further, the total number of equivalent triangle patterns over the integral target is achieved. Combining the TOD criteria, the target acquisition ranges for different discrimination (detection, recognition, and identification) levels are obtained by solving a set of equations. Comparisons with the preliminary experimental results show that this new TA approach can provide reasonable prediction of field performance.     

TOD Performance Theoretical Model for Staring Thermal Imagers

The TOD (triangle orientation discrimination threshold) metric, as a promising sensor performance measure, is suitable for all types of thermal imaging systems and corresponds to the field performance very well. Since this method was proposed, a great progress has been made in some aspects, such as experimental validation, field performance, and TOD sensor performance model. However, what is still missing is a TOD curve prediction model similar to MRTD (minimum resolvable temperature difference) equation. Here, an equation-based TOD performance model was first developed focusing on staring thermal imagers. Specifically, spatial distribution of the triangle standard test pattern is described quantitatively. Resolution transfer characteristic of the staring infrared imagers is modeled with linear system theory, and system noise is characterized with the introduction of three-dimensional noise model. The modulation effect of overall subsystem of staring thermal imager on the non-periodic triangle standard pattern is analyzed. The matched filter idea is adopted to characterize quantitatively the spatial integration of the eye to signal and various noise components over the triangle pattern area, and the perceived signal-to-noise ratio of the staring thermal imager is derived. Then, the TOD performance theoretical model is established. Comparing with experimental results, this theoretical model gives reasonable prediction of the TOD performance curve of staring thermal imagers. Although more tests and modifications are required, these preliminary results suggest that this model can be developed into a model that predicts the TOD for all kinds of sensors.     

TOD Performance Model for Staring Thermal Imager with Machine Vision

A triangle orientation discrimination threshold(TOD) performance theoretical model is derived for the staring thermal imager based on machine vision. Specifically, how to obtain the TOD curve based on machine vision is briefly described. The spatial frequency distribution of the triangle test pattern is first determined. The transform and response characteristics of the non-periodic triangle pattern and its background clutter through machine vision-based thermal imager are analyzed. The three-dimensional matched filter is adopted to characterize quantitatively the spatial and temporal integration of image enhancement algorithms to the output triangle pattern signal, various noise components and background clutter, and the signal-to-interference ratio (SIR) of the triangle pattern output image is derived for the staring thermal imager based on machine vision. Then, the TOD performance theoretical model is established by assuming that the output SIR is equal to the threshold SIR_75% determined by the discrimination criteria of machine vision. Preliminary simulation and experimental results show that this theoretical model can give reasonable prediction of the TOD performance curve for staring thermal imagers based on machine vision.     

表征光电成像系统性能的两种不同方法

比较了表征光电成像系统性能的两种方法———最小可分辨温差法/最小可分辨对比度(MRTD/MRC)法与三角形方向辨别(TOD)法在测试样条、测量方法、适用范围、性能预测精度等方面的差别,分析了这些差别的产生机理,展望了TOD法在未来光电成像系统性能评估和测试领域的应用。     

红外图像转换器传递函数测试方法

红外图像转换器是红外成像制导武器半实物仿真系统红外成像目标模拟器的重要组成部分,用于产生动态红外目标/干扰环境特性图像。图像传递函数是评价图像转换器性能优劣的一项综合性技术指标,包含了图像在不同等效黑体温度条件下,不同分辨率对应的图像调制度。叙述了红外图像转换器图像传递函数的定义和测试方法,用经过标定的热像仪和光学放大系统对红外图像转换器产生的图像进行放大,测试了亮、暗条纹的辐亮度,计算了条纹对的对比度。描述了红外热像仪的标定方法和使用注意事项,给出了256×256元MOS电阻阵列红外图像转换器图像传递函数的测试结果。该文可作为红外图像转换器性能考核的参考。     

Field Performance Prediction and IR Imaging Simulation Based on Measured System Level Parameters*

Traditionally, the Minimum Resolvable Temperature Difference (MRTD) and the Johnson criteria are used to predict the field performance of IR imaging system. However this method generally leads to far too pessimistic range predictions. For the improvement of the prediction accuracy of the field performance, a novel approach to predict field performance is proposed based on the three-dimensional infrared scene generated by Vega software. Further, this approach utilizes the measured system level parameters to characterize the signal transfer process, noise, and the blur effect of the output image instead of theoretical model. By controlling the target range in the simulated image, a simulation experiment is performed, and the range corresponding to the 75% correct probability of discrimination is achieved by the statistical method. Comparisons with the real experimental result show that this method can give more accurate range prediction than the target acquisition (TA) model based on the MRTD.     

Performance analysis of an automated E-pulse target discriminationscheme

An automated E-pulse scheme for target discrimination was initially presented by Ilavarasan et al. (1993) without an analytic performance evaluation. Assuming that target responses are contaminated with white Gaussian noise, an automated E-pulse scheme is rigorously analyzed to yield a reliable measure of performance. The discrimination performance of this automated E-pulse scheme is determined quantitatively through the use of energy discrimination numbers (EDNs). Statistics of the EDNs are evaluated analytically to derive the probability of correct identification. The probability of identification as a function of signal-to-noise ratio (SNR) is evaluated using the theoretical scattering data for all potential targets to predict the performance of the automated E-pulse scheme. These theoretical results are corroborated by direct simulation of the discrimination scheme. In addition, the probability density functions of the EDNs are presented providing new physical insights into E-pulse performance as a function of target geometries and SNR     

Uncertainty Analysis in Triangle Orientation Discrimination Threshold Measurement

The Triangle Orientation Discrimination Threshold (TOD) is a new figure of merit for characterizing the performance of an infrared imager. In this paper, focusing on a reflective infrared target projector, the TOD curve is measured utilizing four alternative forced choice (4AFC) test method. Furthermore, the dominant sources affecting uncertainties in the TOD measure values, are analyzed in detail, the effective temperature difference, sensor, and observer. And, the variation of the TOD measurement is described quantitatively utilizing the uncertainty analysis theory. Simulation results show that the TOD measurement values vary in a small range, which show a higher stability than minimum resolvable temperature difference (MRTD) measure values.