基于边缘检测技术的温度计自动检定装置研制

为了解决人工检定温度计存在的工作效率低、劳动强度高、读数误差大和检定油烟重等问题,运用边缘检测技术,先对图像作预处理,再利用经典的Sobel边缘检测算子和三线识别方法,将温度计液柱图像转换为数字化的温度计示值;对温度计检定槽进行自动化改造,由计算机控制温度计插槽旋转角度和摄像机运动高度,再由单片机控制执行指令,将采集的图像传给计算机,从而研制出温度计自动检定装置,并实现计算机自动处理检定信息、打印原始记录和检定证书。     

基于视频的运动目标检测算法的比较与分析

帧差法和背景差分法是目前常用的基于视频的运动目标检测方法。首先从帧差法和背景差分法中选取了几种具有代表性的算法;然后针对典型场景条件对这些方法进行了实验测试,并利用实验结果对算法进行了定性和定量的比较和分析。实验结果表明,ViBe算法、码本模型、混合高斯模型、自适应混合高斯模型、中值滤波和自适应背景建模具有较好的检测效果,但任何一种方法都有其局限性。     

面元法的红外制导图像边缘检测

红外图像通常存在着噪声大、目标和背景之间具有较小的灰度差、边缘较模糊的特点,使现有的方法不能有效地提取红外图像的边缘。而图像的表面积特征能较好地针对红外图像的这些特征提取出边缘,因此利用图像的表面积特征提出了一种实用的面元边缘检测法。在对图像灰度值变化情况进行分析的基础上,使用了一种十字面积计算公式以满足边缘检测的需要。通过噪声误差对面元法和梯度法计算结果的影响分别进行分析,发现面元法比梯度法的噪声抑制能力至少高两倍,而且在边缘处的噪声抑制能力更高。将面元法几种常用模板算子法的红外图像边缘检测结果进行实际对比后发现,所提出的面元法对红外图像的目标边缘检测能取得良好的结果。     

基于Hough变换的高分辨SAR图像道路目标检测

给出了一种检测高分辨率SAR图像道路目标的方法。该方法首先对原始图像进行一系列的预处理,然后利用Hough变换识别道路。在提取感兴趣区域时,该算法采用了高斯概率迭代方法代替了一般的CFAR检测或阈值分割;在进行Hough变换时,该算法用平均Hough变换代替一般的Hough变换;在检测变换域峰值点时,采用了结合全局CFAR检测的最大值点搜索的方法;最后对检测出来的准道路,该算法根据道路的独具性质进行鉴别,以确定其是否确实为道路。使用MSTARRedstone实测杂波图进行实验,取得了满意的效果。     

自然场景车标数据集的构建及其应用

车标作为车辆身份的关键特征之一,在车辆的监控与辨识中发挥着重要作用。由于自然场景复杂多变,对其中的车标进行准确识别仍具有很大的挑战性。目前公开数据库很少且存在诸多局限,导致研究缺乏可信度和实用性。本文建立了一个面向自然场景的全新数据集,包含多种采集环境下的10 324幅、67类车辆图像。基于此数据集开展应用研究,提出一个目标检测与深度学习相结合的车标识别方法,包括车标区域定位和车标种类预测两大步骤。实验表明,该方法对复杂背景有较强的适应性,在涉及30种车标的分类任务中达到89.0%的总体识别率。     

Bidirectional parallel multi-branch convolution feature pyramid network for target detection in aerial images of swarm UAVs

In this paper, based on a bidirectional parallel multi-branch feature pyramid network (BPMFPN), a novel one-stage object detector called BPMFPN Det is proposed for real-time detection of ground multi-scale targets by swarm unmanned aerial vehicles (UAVs). First, the bidirectional parallel multi-branch convolution modules are used to construct the feature pyramid to enhance the feature expression abilities of different scale feature layers. Next, the feature pyramid is integrated into the single-stage object detection framework to ensure real-time performance. In order to validate the effectiveness of the proposed algorithm, experiments are conducted on four datasets. For the PASCAL VOC dataset, the proposed algorithm achieves the mean average precision (mAP) of 85.4 on the VOC 2007 test set. With regard to the detection in optical remote sensing (DIOR) dataset, the proposed algorithm achieves 73.9 mAP. For vehicle detection in aerial imagery (VEDAI) dataset, the detection accuracy of small land vehicle (slv) targets reaches 97.4 mAP. For unmanned aerial vehicle detection and tracking (UAVDT) dataset, the proposed BPMFPN Det achieves the mAP of 48.75. Compared with the previous state-of-the-art methods, the results obtained by the proposed algorithm are more competitive. The experimental results demonstrate that the proposed algorithm can effectively solve the problem of real-time detection of ground multi-scale targets in aerial images of swarm UAVs.     

Electronic warfare in the optical band: Main features,examples and selected measurement data

The paper presents the possibilities of, and methods for, acquiring, analysing and processing optical signals in order to recognise, identify and counteract threats on the contemporary battleground. The main ways electronic warfare is waged in the optical band of the electromagnetic wave spectrum have been formulated, including the acquisition of optical emitter signatures, as well as ultraviolet (UV) and thermal (IR) signatures. The physical parameters and values describing the emission of laser radiation are discussed, including their importance in terms of creating optical signatures. Moreover, it has been shown that in the transformation of optical signals into signatures, only their spectral and temporal parameters can be applied. This was confirmed in experimental part of the paper, which includes our own measurements of spectral and temporal emission characteristics for three types of binocular laser rangefinders. It has been further shown that through simple registration and quick analysis involving comparison of emission time parameters in the case of UV signatures in “solar-blind” band, various events can be identified quickly and faultlessly. The same is true for IR signatures, where the amplitudes of the recorded signal for several wavelengths are compared. This was confirmed experimentally for UV signatures by registering and then analyzing signals from several events during military exercises at a training ground, namely Rocket Propelled Grenade (RPG) launches and explosions after hitting targets, trinitrotoluene (TNT) explosions, firing armour-piercing, fin-stabilised, discarding sabots (APFSDS) or high explosive (HE) projectiles. The final section describes a proposed model database of emitters, created as a result of analysing and transforming the recorded signals into optical signatures.     

基于RBF神经网络的背景估计及红外小目标检测

利用RBF神经网络的函数逼近特性,得到了一种红外图像背景估计算法,进而提出了一种检测红外小目标的方法。利用有目标和没有目标的真实红外图像对此算法进行检测,背景估计效果理想,小目标检测效果理想,证明该算法是可行有效的。     

自修复飞控系统故障检测与隔离软件系统

介绍了一个自修复飞行控制系统故障检测和隔离实时工程软件系统,该系统按照软件工程规范、采用等价空间法和残差向量的概率信息融合法研制和开发,不仅能对飞行控制系统的突变故障——作动器卡死故障和控制舵面损伤故障进行实时检测和隔离,而且具有良好的用户界面。仿真研究表明该软件系统具有良好的鲁棒性和实时性。     

低可观测条件下的一种信号检测方法

针对在低信噪比条件下,检测机载合成孔径雷达(SAR)信号比较困难的情况,综合谱图-Hough变换法可以进行能量积累和小波变换具有比较有效抑制杂波的特点,提出了一种检测SAR信号的方法,即基于谱图-Hough变换和小波变换的检测方法。运用该方法首先对SAR信号进行谱图-Hough变换,谱图-Hough变换能对SAR信号进行能量积累,降低虚警概率,然后使用小波变换排除Hough空间中的虚假峰值。仿真结果表明了该算法能够有效地抑制杂波,达到检测的目的。