融合灰度与显著性特征的空中红外目标跟踪

针对红外目标特征简单且信息量少导致跟踪精度不高,提出一种基于灰度和显著性特征融合的核相关滤波算法用于空中红外目标跟踪。首先,在保证目标足够特征信息量的前提下对较大的目标进行不同等级压缩。然后将提取的二维灰度特征与显著性特征按页方式进行拼接扩展成三维特征,再将融合的特征用于核相关滤波。实验表明所提算法能够适应多种环境下的空中红外目标跟踪,跟踪精度和成功率典型值分别达到84.8%和63.9%,较大部分算法有很大提高,平均跟踪速度高达125 f/s,体现出了良好的实时性。因此,本文提出的算法在保证实时性的同时提高了跟踪的可靠性,具有一定的实用意义。

Field study of seasonal thermal comfort and adaptive behavior for occupants in residential buildings of Xi’an,China

The study aims to investigate the thermal comfort requirements in residential buildings and to establish an adaptive thermal comfort model in the cold zone of China.A year-long field study was conducted in residential buildings in Xi’an,China.A total of 2069 valid questionnaires,along with indoor environmental parameters were obtained.The results indicated occupants’thermal comfort requirements varied with seasons.The neutral temperatures were 17.9,26.1(highest),25.2,and 17.4℃(lowest),and preferred temperatures were 23.2,25.6(highest),24.8,and 22.4℃(lowest),respectively for spring,summer,autumn,and winter.The neutral temperature and preferred temperature in autumn are close to the neutral temperature in summer,while the neutral temperature and preferred temperature in spring are close to that in winter.Besides,the 80%and 90%acceptable temperature ranges,adaptive thermal comfort models,and thermal comfort zones for each season were established.Human’s adaptability is related to his/her thermal experience of the current season and the previous season.Therefore,compared with the traditional year-round adaptive thermal comfort model,seasonal models can better reflect seasonal variations of human adaptation.This study provides fundamental knowledge of the thermal comfort demand for people in this region.

采用信息融合的红外目标检测算法

针对空中红外目标检测过程中存在的检测精度不高、易受干扰等问题,本文提出了一种信息融合的目标检测算法。首先,针对单一滤波模式的不足,采用降采样方式将多种模式的滤波方法进行融合,有效提高了滤波后的图像信噪比,减少了预处理时间。然后,对质心检测选定的检测区域进行显著性检测和边缘检测,并将结果进行融合,增强了目标区域的信号强度,改善了算法的检测精度。文中从理论上介绍和分析了该算法对红外目标检测的有效性,并通过Matlab仿真实验与其他单一检测算法进行了对比。实验结果表明,信息融合的检测算法能够在多种复杂情况下有效检测出空中红外目标,证明融合检测算法的抗干扰性和适应性上更强、检测精度更高、算法的鲁棒性更好。