结合超分重建的室内指示牌检测方法

打开文本图片集

Indoor Signage Detection Method with Super-resolution Reconstruction

YUAN Fei1 RAO Raol ZHOU Guanghai² CHEN Weibin1 （1.School of Automation， Guangdong Polytechnic Normal University， Guangzhou ， China 2.Guangzhou Southern Surveying and Mapping Technology Co.， Ltd.， Guangzhou ， China）

Abstract：To adressthe low detection acuracyof distant indoor signage，this paper proposes an indoor signage detection method with super-resolutionreconstruction.Analysis ofthe monocularcamera perspective imaging modelreveals that the imaging pixels ofsignagederease withincreasingshootingdistance.Byapplyingsuper-resolutionreconstruction toimages containing indr signage，theresolutionofsignageimagesisanced，threbyimproingdetectionacuracy.ComparativeexperimentsonteHN dataset and a laboratory dataset demonstrate that images reconstructed at 2× and 4× super-resolution using SRCNN， SRResNet， and SRGANsuper-resolutionreconstructionneuralnetwork modelsignificantlyoutperformon-reconstructedimages insignagedetection accuracy.

Keywords： super-resolution reconstruction; indoor signage detection;camera perspective imaging model; signage imaging calculation method; object detection algorithms

0 引言

基于位置信息的研究主要通过获取目标物的位置信息来确定其具体位置，在博物馆、机场、超市、医院和地下矿井等大型复杂室内环境中应用广泛[1-2]。（剩余8506字）