一种改进的Retinex彩色图像增强方法

传统Retinex算法在对彩色图像进行增强时,容易产生光晕以及处理后图像整体对比度较低等问题.为此提出了一种改进Retinex的图像增强方法,该方法采用三高斯模型和高斯滤波相结合的双边滤波对照度分量进行估计;然后在提取反射分量时,引入原始图像亮度影响因子实现区域自适应增强;最后进行归一化调整和彩色图像恢复,实现图像增强.实验结果表明,该方法能够在增强图像的同时,有效抑制光晕,提高图像的整体对比度.     

多特征多阈值级联AdaBoost行人检测器

为了满足更快、更准、更鲁棒的行人检测需求,考虑交通监控视频图像质量不高与局部特征不明显的缺点,采用简单的行人特征来实现行人检测。除矩形度、高宽比、轮廓复杂度、宽度比、行人面积特征外,特定选用了对遮挡等干扰具有强鲁棒性的头部圆形度这一简单的局部特征。考虑交通监控视频图像中行人的尺寸变化,引入区域划分策略划分图像区域。兼顾高检测率和低误检率,根据分类误差最小原则与正样本分类率最大原则训练多个单特征多阈值AdaBoost行人检测器。为了优化多个行人检测器级联后的检测性能,在兼顾检测性能和检测速度的基础上,定义了以贡献率作为行人检测器的级联规则,依据贡献率大小确定的级联次序为基于高宽比、宽度比、矩形度、行人面积、轮廓复杂度和头部圆形度的行人检测器,依次进行级联,建立了新的多特征多阈值级联AdaBoost行人检测器。选用3个交通场景对行人检测器进行测试,并与单级AdaBoost行人检测器与现有2种级联AdaBoost行人检测器进行比较。分析结果表明:在3个交通场景的检测中,相比其他几种行人检测器,多特征多阈值级联AdaBoost行人检测器具有较高检测率、较快的检测速度和较低误检率,检测率最低为96.70%,误检率最高为0.67%,检测时间小于5s,满足交通场景中对行人检测实时性和可靠性的要求。     

基于改进Canny算子的遥感图像边缘检测

针对传统的Canny边缘检测算子存在高斯滤波函数方差和阈值需要人工设置的缺点,提出了一种基于改进Canny算子的遥感图像边缘检测算法.首先利用复合形态学滤波代替高斯滤波,可得到更好的平滑效果,同时保留更多的边缘信息;然后利用Otsu阈值方法对阈值进行自适应设定,可以使检测边缘更加连续完整,并减少虚假边缘的存在;最后引入数学形态学算法实现对边缘细化处理.实验结果表明,将改进的Canny算子应用到遥感图像的处理中,具有良好的抗噪性能和检测精度.     

一种新型图像匹配方法在视觉导引 AUV 对接中的应用

针对视觉导引AUV对接过程图像的变化特点,提出一种量子粒子群算法与灰度关联分析方法相结合的图像匹配算法。此算法结合量子粒子群并行搜索的快速性与灰色关联分析方法较强的鲁棒性,对旋转、平移、亮度变化等不敏感,符合视觉导引AUV对接过程中对图像匹配算法的要求。详细介绍了基于量子粒子群优化算法的图像匹配算法,算法以图像灰度直方图的灰色绝对关联度作为适应度函数。利用水池实验所得图像进行测试,实验结果显示,所提出的图像匹配算法准确、快速、鲁棒性强,能够很好地应用于视觉导引AUV对接。     

基于船舶卫星遥感图像的目标特征算法

在现代海洋运输及海上军事业务中,对船舶目标的有效定位及跟踪是保证海洋运输安全的有效方法之一。现有的海上目标检测跟踪系统中,有基于雷达目标检测﹑水声目标检测及卫星遥感图像的目标检测方法。由于海上环境的复杂性及目标物的快速移动性,基于遥感图像的目标检测成为最重要的方法之一。本文分析现有船舶目标物的特征参数,提出船舶卫星遥感图像快速有效的船舶目标特征值提取算法,最后给出了基于此算法的数据库设计。     

SMA沥青混凝土路面材料荷载行为模式研究

采用图像处理分析技术观察SMA沥青混凝土及密级配沥青混凝土在荷载作用下其粗集料移动轨迹的变化,定量描述SMA沥青混凝土粗集料构架形成机制、抗车辙能力成因、粗集料材料形状组合因子偏差等因素,建立粗集料移动轨迹与车辙量的行为模式,研究SMA沥青混凝土材料荷载行为,以使其应用于公路及机场路面能完全发挥其特性。研究结果表明:SMA沥青混凝土粗集料碾压过程中粗集料趋向水平方向移动,且移动量小,粗集料构架处于最稳定阶段,故SMA沥青混凝土车辙量较小,抗车辙能力较好;密级配沥青混凝土中粗集料趋向垂直方向移动,粗集料移动量大,容易造成粗集料构架不稳定,故沥青混凝土车辙量较大。     

An innovative method for the analysis of vehicle movements in roundabouts based on image processing

The objective of this paper is to propose a method, based on the image processing of field survey data, to analyze vehicles movements into roundabouts. This research study consisted of three stages: a field survey to collect vehicular flow images captured by video cameras, the processing of these images using a proprietary software (VeTRA—Vehicle Tracking for Roundabout Analysis), and finally, the analysis of the collected data. The main feature of the software is that it allows the automatic computation of the main variables necessary to rank and evaluate a generic roundabout: the entry/exit (E/E) matrix with classification of vehicles (e.g., heavy, light, and motorbikes), vehicle trajectories, and vehicular speed diagrams along the paths inside the roundabout. The processing system is robust enough to withstand classic problems affecting image processing such as variable wind conditions, cloud cover, shadows, and obstructions. Calibration and error evaluation have been deduced from data collected by a high precision Real Time Kinematic GPS video recording system mounted on a probe vehicle. Data of E/E matrices generated by VeTRA are compared with those manually counted on the corresponding video images. A case study of an existing roundabout is featured in the paper. The results indicate that the software has a high capability of generating the E/E matrix. The analysis of vehicular trajectories with both the plot of curvature diagrams and the corresponding speed diagrams enable the evaluation of driver behavior relative to the geometric shape of the roundabout. Copyright © 2011 John Wiley & Sons, Ltd.     

Differences in air passengers’ buying behaviour: findings from Korean and Australian international passengers

Abstract

This paper investigates how air passengers’ expectations, ticket price, airline service quality, value, passenger satisfaction and airline image determine their buying behaviour. To test the conceptual frameworks, path analysis was applied to data collected from Korean and Australian international passengers to examine differences between these two groups. Further analyses were undertaken on different passenger segments between national and foreign airline passengers. The results of the path analysis reveal that air passengers’ buying behaviour differs significantly between Korean and Australian international passengers. Results also reveal that the determinants of air passengers’ buying behaviour differ by airlines.     

基于遥感图像信息特征的单调递增SSDA算法

图像配准就是将不同时间、不同传感器或不同条件下对同一景物获取的两幅或多幅图像,进行比较找到该组图像中的共有景物,或是根据已知模式到另一幅图中寻找相应的模式。利用遥感图像进行目标监测需要进行图像配准处理。匹配算法如何达到高精度、高匹配正确率和实时性成为人们追求的目标。文章在传统匹配算法的基础上,提出两点改进:一是通过PCA-圆形SIFT算法提取图像特征角点,降低维数,优化计算;二是利用图像角点作为单调递增阈值序列SSDA算法匹配的基本像素点,利用遥感图像信息特征降低匹配计算量。最后进行实验仿真,结果表明,改进后的算法使得配准速度进一步提高,能够满足遥感图像配准实时性的要求。     

Image analysis of aggregate, mastic and air void phases for asphalt mixture

The shape characterization and spatial distribution of aggregate,mastic and air void phases for asphalt mixture were analyzed.Three air void percentage asphalt mixtures,4%,7% and 8%,respectively,were cut into cross sections and polished.X-ray scanning microscope was used to capture aggregate,mastic,air void phase by the image.The average of polygon diameter was chosen as a threshold to determine which aggregates would be retained on a given sieve.The aggregate morphological image from scanned image was utilized by digital image processing method to calculate the gradation of aggregate and simulate the real gradation.Analysis result shows that the air void of asphalt mixture has influence on the correlation between calculation gradation and actual gradation.When comparing 4.75 mm sieve size of 4%,7% and 8% air void asphalt mixtures,7% air void asphalt mixture has 55% higher than actual size gradation,8% air void asphalt mixture has 8% higher than actual size gradation,and 4% air void asphalt mixture has 3.71% lower than actual size gradation.4% air void asphalt mixture has the best correlation between calculation gradation and actual gradation comparing to other specimens.The air void percentage of asphalt mixture has no obvious influence on the air void orientation,and three asphalt mixtures show the similar air orientation along the same direction.