With the advancement of unmanned aerial vehicle(UAV)remote sensing technology,the application of UAVs equipped with thermal infrared lenses has gradually expanded,and the quality assessment of UAV thermal infrared orthophotos has grown increas-ingly crucial.This research aims to optimize the stitching process of UAV thermal infrared images to fulfill the application requirements of various fields,particularly in guiding geothermal prospecting through the detection of surface thermal anomalies by UAVs.The study employs both subjective and objective methods for evaluation.By extracting information entropy,variance,standard deviation,and tem-perature values,and calculating the average gradient of the images,the quality of thermal infrared images processed by different stitc-hing software for the same flight mission is compared.The study area is located in Jianjie Village,Ganzhou,Jiangxi Province.A DJI UAV equipped with FLIR thermal infrared lens was used to conduct flight operations in the delineated area,with a focus on comparing the performance of Pix4D and PhotoScan in thermal infrared image stitching.The results indicate that PhotoScan outperforms Pix4D in terms of the retention of image information and the clarity after stitching,and its thermal infrared images are closer to the originally ac-quired temperature values.Conversely,Pix4D demonstrates superior performance in image completeness and overall coordination,mak-ing it more applicable for the selection of target areas in geothermal prospecting projects.
关键词
无人机/热红外遥感/影像拼接/质量评价/平均梯度/江西赣州
Key words
UAV/thermal infrared remote sensing/image stitching/quality evaluation/mean grade/Ganzhou in Jiangxi Province
维普
万方数据