Combined Filtering Method for Offshore Oil and Gas Platform Point Cloud Data Based on KNN_PCF and Hy_WHF and Its...

Area of Science:

• Marine Engineering
• Geospatial Data Processing
• Robotics and Automation

Background:

• Deep-sea oil exploration requires precise assessment and maintenance of marine structures.
• Traditional measurement methods struggle with the scale and complexity of offshore platforms.
• Three-dimensional (3D) laser scanning offers potential but faces challenges with marine environmental noise.

Purpose of the Study:

• To develop an effective method for processing noisy 3D laser scan data from offshore oil and gas platforms.
• To improve the accuracy and efficiency of 3D data acquisition for marine structures.
• To provide reliable dimensional data for prefabrication and maintenance.

Main Methods:

• A novel filtering approach combining K-Nearest-Neighborhood (KNN) filtering with hyperbolic function-based weighted hybrid filtering.
• Application of the algorithm to process point cloud data acquired from offshore oil and gas platforms.
• Comparative analysis against existing technologies to evaluate performance.

Main Results:

• The proposed method significantly improves noise point filtering efficiency by approximately 11%.
• The total error in data processing is reduced by 0.6 percentage points compared to current methods.
• The algorithm effectively removes noise and preserves critical structural details, even in high-density areas.

Conclusions:

• The developed filtering technique offers superior performance for 3D laser scan data processing in challenging marine environments.
• This method enhances data accuracy and optimizes the 3D reconstruction of offshore platforms.
• The reliable dimensional information generated supports efficient land-based prefabrication of marine structures.