摘要
提出基于图像显著度和信息量均衡的地标链生成方法,该方法建立地标显著度、全景图信息量和地标切换次数的多目标模型,基于遗传算法生成多目标优化的地标链。实验表明,能较好地平衡地标显著度、全景图信息量及地标切换次数等优化目标,生成的地标链具有较高的显著度与较低的视觉认知负担。
The landmark sequence can provide a clear navigation clue in the process of pedestrian navigation. This paper proposes a method of the landmark sequence generation which based on the balance of image significance and information. This method builds a multi-objective model by the information of the panorama, the number of landmark switch and the landmark significance. And based on the genetic algorithm, the multi-objective optimization of the landmark sequence generation method is realized. Experiment shows that this method can balance the three optimization targets well: the landmark significance, the information of the panorama and the number of the landmark switch. Compared with other methods of the landmark sequence generation, this method has a high degree of the landmark significance, the low visual cognitive burden, and can reduce the number of landmark switch.
作者
倪文强
方志祥
李灵
NI Wenqiang;FANG Zhixiang;LI Ling(State Key Laboratory of Information Engineering in Surveying Mapping and Remote Sensing,Wuhan University,Wuhan 430079,China)
出处
《测绘地理信息》
2019年第6期6-10,共5页
Journal of Geomatics
基金
国家自然科学基金(41371420,41771473)
关键词
行人导航
地标链
视觉显著性
图像信息量
遗传算法
pedestrian navigation
landmark sequences
visual significance
image information
genetic algorithm(GA)
