基于VIIRS数据的油气平台提取技术研究

油气平台作为海上油气资源勘探开发的主要设备之一,其数量与空间分布反映了一个区域油气资源的开发状况。普通光学影像易受天气状况影响,而雷达数据成本较高,这给油气平台的检测带来了一定的困难。由于油气平台作业过程中,需要灯光照明,同时伴气的燃烧也产生很强的灯光,因此可以通过检测灯光来实现油气平台的提取。针对VIIRS数据具有强夜间光探测能力,本文提出了一种卷积运算临界值法对海上油气平台进行提取。首先对2期不同时相的VIIRS数据进行卷积运算,对亮像元进行了增强处理,对背景像元进行弱化处理,从而明确区分疑似目标与背景,然后以0值为分界点,对疑似目标进行提取,最后利用油气平台相对静止的特性,通过邻域分析,实现油气平台的提取。结果表明,本文提出的卷积运算临界值法可以有效地提取油气平台,提取准确率约为85.4%,同时可以有效地减少由于经验阈值对提取结果造成的误差。

Research on Oil and Gas Platform Extraction Technology Based on VIIRS Data

As one of the major equipment for offshore oil and gas exploration, the number and spatial distribution of oil and gas platform would reflect the situation of exploitation in a region. Nowadays, it is difficult to acquire data of oil and gas platform because the traditional field observation method is expensive and time-consuming. With the development of remote sensing technology, remote sensing images as an important data source are applied in the extraction of marine targets effectively. However, traditional optical images are easily influenced by cloud and the radar data are too expensive, which make oil and gas platform detection difficult. During the operation process, the oil and gas platform need lamplight, and gas combustion also generate strong light. Thus, the oil and gas platform extraction can be done by detecting light. VIIRS data has a strong ability of detecting light at night and it can be used for the data extraction of oil and gas platform. Considering the feature of VIIRS data, this paper proposed a convolution operation threshold algorithm to extract offshore oil and gas platform and chose Pearl River Mouth basin as a study area for experiment. Firstly, VIIRS data of different time phase were operated to enhance the pixel value and weaken the background pixels so that the suspected target and background can be distinguished. Then, we took the 0 value as the dividing point to extract suspected targets. Finally, we extracted oil and gas platform through the neighborhood analysis making use of the relatively static characteristic of oil and gas platform. The results showed that the accuracy of extracting oil and gas platforms was about 85.4% and the method could effectively reduce the impacts of empirical threshold. At last, this study analyzed the shortage and causes of using VIIRS data to extract oil and gas platforms.