A new method of splicing infrared thermal Unmanned aerial vehicle(UAV) image with special acquisition mode is presented. We used the bi-directional whiskbroom scanning frame infrared UAV (300,000 pixels) developed by the Shanghai Institute of Technical Physics,Chinese Academy of Sciences(SITP) to obtain the image of an area in haiyan,Zhejiang province.In order to get more accurate temperature data in a wide range, correction and optimization of the mosaic strategy of this infrared image are needed.In this paper, an orthographical correction model based on drone position and orientation system (POS) parameters was established, By optimizing scale invariant feature transform (SIFT) matching parameters and Random Sampling Consensus (RANSAC) algorithm, more reliable matching results were obtained.After rough calculation of adjacent images by image location to calculate whether it is an adjacent image, this can reduce the operation time of the splicing algorithm.The overlapping area images fused with the multi-resolution pyramid algorithm. Finally, the large area image and temperature inversion map of the study area were obtained. Inversion results showed that the error of temperature inversion less than 0.2 degree by comparing with the original temperature before unspliced.It can meet the subsequent application requirements of the UAV infrared image.

Date Published: 12 March 2019
PDF: 7 pages
Proc. SPIE 11023, Fifth Symposium on Novel Optoelectronic Detection Technology and Application, 1102336 (12 March 2019); doi: 10.1117/12.2521929

Published in SPIE Proceedings Vol. 11023:
Fifth Symposium on Novel Optoelectronic Detection Technology and Application
Qifeng Yu; Wei Huang; You He, Editor(s)