Estimating the Volume of Oil Tanks Based on High-Resolution Remote Sensing Images

Tong Wang, Ying Li, Shengtao Yu, Yu LiuRemote 2019 · DOI

TL;DR

This paper presents a novel method for estimating the volume of oil tanks using high-resolution satellite imagery. By analyzing the crescent-shaped shadows of oil tanks with a sub-pixel subdivision technique and employing the Hough transform for radius detection, the researchers achieve high precision in volume estimation, crucial for global oil storage monitoring.

What Problem Does This Paper Address?

Accurately estimating the volume of oil tanks is vital for global oil reserve management, economic planning, and transportation logistics. While remote sensing offers a way to monitor these assets, traditional methods for extracting information from satellite images often struggle with the unique characteristics of oil tank shadows. Unlike the rectangular shadows of buildings, oil tank shadows are crescent-shaped, making standard shadow length calculation difficult. This paper aims to bridge this gap by developing a robust method to accurately extract oil tank volumes from high-resolution satellite images, addressing the need for precise global oil storage capacity assessment.

Methods

The proposed method involves several key steps:

  • Image Preprocessing: Standard procedures like radiation correction, geometric correction, image fusion, and cropping are applied to prepare the satellite imagery.
  • Shadow Extraction: The core of the shadow detection lies in converting the image to the HSV (Hue, Saturation, Value) color space. This is followed by Otsu's automatic thresholding, shadow area thresholding, and morphological closing operations to isolate the oil tank shadows. The HSV space is chosen because shadows exhibit distinct hue and value characteristics.
  • Shadow Length Calculation: A novel "median method based on sub-pixel subdivision positioning" is introduced to accurately measure the length of the crescent-shaped tank shadows. This method refines traditional approaches by dividing shadow segments into smaller units for more precise measurement.
  • Tank Height Calculation: The height of the oil tank is determined using the calculated shadow length and geometric relationships involving solar elevation angle, solar azimuth, and satellite parameters.
  • Tank Radius Calculation: The Hough transform, specifically the "HoughCircles()" function, is employed to detect the circular top of the oil tank and determine its radius.
  • Volume Calculation: Finally, the oil tank volume is calculated using the derived height and radius, assuming a cylindrical shape.

Datasets

The study utilizes high-resolution optical remote sensing images from the Gaofen 2 satellite. This satellite provides panchromatic imagery with a spatial resolution of 0.81 m and multispectral imagery with a resolution of 3.20 m. The experimental data consists of images of oil tanks located in a national oil reserve base.

Key Insights and Ideas

  • Crescent Shadow Challenge: The paper highlights that oil tank shadows are crescent-shaped, posing a unique challenge for traditional shadow analysis methods designed for rectangular shapes.
  • HSV Color Space Advantage: Utilizing the HSV color space effectively enhances shadow detection by leveraging distinct hue and value properties of shadowed regions.
  • Sub-pixel Precision for Shadows: The proposed sub-pixel subdivision positioning method is crucial for accurately measuring the length of these irregular crescent shadows, leading to more precise height estimations.
  • Hough Transform for Circular Objects: The application of the Hough transform is a well-established technique for detecting circular shapes, and its use here effectively identifies the tank's top radius.
  • Median Method Superiority: The median method for shadow length calculation, when combined with sub-pixel subdivision, demonstrates superior accuracy compared to pixel number-based or direct measurement methods.

Key Results

  • Shadow Extraction Accuracy: The proposed shadow extraction method achieved a precision of 96.05%, outperforming traditional histogram threshold and morphological methods.
  • Tank Height Accuracy: The median method for shadow length calculation resulted in a mean absolute error of 0.238 m for tank height, with a relative error within 1.15%. This is noted as being better than previous work.
  • Tank Volume Accuracy: The absolute error for calculated tank volumes ranged from 416 to 3050 m³, and the relative error ranged from 0.38% to 2.78%. The estimation error for most tank volumes was within 3000 m³.

Limitations and Future Work

  • Hough Transform Limitations: The authors acknowledge that the Hough transform did not identify all oil tanks in the study.
  • Future Work:
    • Investigate the nature of shadows in other color spaces and incorporate texture features for shadow extraction from a wider range of images.
  • Combine the Hough transform with machine-learning algorithms to improve the accuracy of tank identification and radius estimation.

Summary updated

Abstract

The purpose of this study is to obtain oil tank volumes from high-resolution satellite imagery to meet the need to measure oil tank volume globally. A preprocessed remote sensing HSV image is used to extract the shadow of the oil tank by Otsu thresholding, shadow area thresholding, and morphological closing. The oil tank shadow is crescent-shaped. Hence, a median method based on sub-pixel subdivision positioning is used to calculate the shadow length of the oil tank and then determine its height with high precision. The top of the tank and its radius in the image are identified using the Hough transform. The final tank volume is calculated using its height and radius. A high-resolution Gaofen 2 optical remote sensing image is used to evaluate the proposed method. The actual height and volume of the tank we tested were 21.8 m and 109,532 m3. The experimental results show that the mean absolute error of the height of the tank calculated by the median method is 0.238 m, the relative error is within 1.15%, and the RMES is 0.23. The result is better than the previous work. The absolute error between the calculated and the actual tank volumes ranges between 416 and 3050 m3, and the relative error ranges between 0.38% and 2.78%. These results indicate that the proposed method can calculate the volume of oil tanks with high precision and sufficient accuracy for practical applications.
BibTeX
@inproceedings{wang2019estimating,
  author = {Tong Wang and Ying Li and Shengtao Yu and Yu Liu},
  title = {{Estimating the Volume of Oil Tanks Based on High-Resolution Remote Sensing Images}},
  booktitle = {Remote},
  year = {2019},
  doi = {10.3390/rs11070793},
}
Figure