Three dimensional generative pre-trained transformer for subsurface imaging and display and methods of use

A system for subsurface imaging and display of a subsurface below a surface, comprising:

• a memory device storing instructions;
• a processor in communication with said memory device, said processor configured to execute said instructions;
• a subsurface image capture module in communication with said processor, said subsurface image capture module having one or more wave generating devices and one or more sensors affixed to a vehicle, configured to capture a series of digital image datasets of the subsurface below the surface, each said digital image dataset including coordinate reference data;
• a Three Dimensional Generative Pre-trained Transformer (3DGPT) model configured to receive a single two-dimensional (2D) RGB image from said series of digital image datasets and generate a 3D depth map via monocular depth estimation, wherein said 3DGPT model comprises an encoder-decoder neural network trained on proprietary datasets including paired RGB images and ground truth depth maps with at least one of a key subject, a foreground, and a background element;
• wherein said processor executes instructions to generate a digital model of said series of digital image datasets while maintaining said coordinate reference data, determine said 3D depth map of said digital model using said 3DGPT model, and identify a key subject point in said digital model;
• and
• wherein said digital image datasets are selected from the group consisting of satellite, ground penetrating radar, LIDAR, topographical, hydrological, sinkhole susceptibility maps, USGS, USACE, and NOAA data and combinations thereof.