A sea spill on-line and fast identification method of oil spill based on unmanned aerial vehicle

• Main Authors: WANG PENGQIAN, PEI WEI, LI FEI, SONG SHILIN, WANG DEGAO, LU MINGYU
• Format: Patent
• Language: Chinese; English
• Published: 01.02.2019
• Subjects: CALCULATING; COMPUTING; COUNTING; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The invention provides a sea spill on-line fast identification method based on an unmanned aerial vehicle, including carrying thermal infrared camera on UAV to patrol the target sea area and collecting the video of observation area in real time. The thermal infrared image of sea surface was obtained from the video and the standard format image was obtained by preprocessing. The maximum stable extreme value region (MSER) algorithm was used to detect the oil film hot spots in the image and the integrated channel feature (ICF) descriptor was used to extract the oil film features. The support vector machine (SVM) classifier is used to train the oil film characteristics, and the oil spill recognition model is established. The oil spill identification model is used to identify the oil spill on the sea. The method of the invention has the advantages of simple steps, high speed and efficiency, high recognition rate, good robustness and strong applicability. 本发明提供种基于无人机的海上溢油在线快速识别方法，包括：无人机搭载热红外相机巡视目标海域并实时采集观测区视频，从视频中获