Generating training samples via 3D modeling for greenhouse gas emission detection
A computer-implemented method for generating training samples via 3D modeling for machine learning-based greenhouse gas emission detection is described. The method includes generating at least one 3D model corresponding to a real-world facility and at least one 3D model corresponding to gas plume emissions and building a virtual environment comprising a 3D field based on the at least one 3D model corresponding to the real-world facility and the at least one 3D model corresponding to gas plume emissions. The method also includes capturing images or video of the 3D field via simulation of the virtual environment and assigning labels to objects in the images or videos.
1. A computer-implemented method for generating training data using three dimensional (3D) modeling for machine learning-based greenhouse gas emission detection in oil and gas fields, the method comprising:
generating, using at least one hardware processor, at least one 3D model corresponding to a real-world facility and at least one 3D model corresponding to gas emission plumes;
building, using at least one hardware processor, a virtual environment comprising a 3D field based on the at least one 3D model corresponding to the real-world facility and the at least one 3D model corresponding to gas emission plumes;
capturing, using the at least one hardware processor, images or video of the 3D field via simulation of the virtual environment;
assigning, using the at least one hardware processor, labels to objects in the images or videos to generate labeled images or videos; and
training, using the at least one hardware processor, machine learning models using training samples including the labeled images or videos.
2. The computer implemented method of claim 1 , wherein simulation of the virtual environment comprises simulation of the 3D field via drone based image capture at a predetermined angle of image capture.
3. The computer implemented method of claim 1 , wherein simulation of the virtual environment comprises simulation of the 3D field via drone based video capture along as the drone traverses a predetermined path.
4. The computer implemented method of claim 1 , wherein the images or video of the 3D field comprise varying spectral bands.
5. The computer implemented method of claim 1 , wherein the machine learning models are trained to detect spectral variations among objects and greenhouse gases in the captured images or video to predict greenhouse gas emissions.
6. The computer implemented method of claim 1 , wherein the at least one a real-world facility corresponds to a physical oil and gas field.
7. The computer implemented method of claim 1 , wherein the gas emission plumes correspond to average emission values from respective oil and gas facilities.
8. An apparatus comprising a non-transitory, computer readable, storage medium that stores instructions that, when executed by at least one processor, cause the at least one processor to perform operations comprising:
generating at least one 3D model corresponding to a real-world facility and at least one 3D model corresponding to gas emission plumes;
building a virtual environment comprising a 3D field based on the at least one 3D model corresponding to the real-world facility and the at least one 3D model corresponding to gas emission plumes;
capturing images or video of the 3D field via simulation of the virtual environment;
assigning labels to objects in the images or videos to generate labeled images or videos; and
training machine learning models using training samples including the labeled images or videos.
9. The apparatus of claim 8 , wherein simulation of the virtual environment comprises simulation of the 3D field via drone based image capture at a predetermined angle of image capture.
10. The apparatus of claim 8 , wherein simulation of the virtual environment comprises simulation of the 3D field via drone based video capture along as the drone traverses a predetermined path.
11. The apparatus of claim 8 , wherein the images or video of the 3D field comprise varying spectral bands.
12. The apparatus of claim 8 , wherein the machine learning models are trained to detect spectral variations among objects and greenhouse gases in the captured images or video to predict greenhouse gas emissions.
13. The apparatus of claim 8 , wherein the at least one a real-world facility corresponds to a physical oil and gas field.
14. The apparatus of claim 8 , wherein the gas emission plumes correspond to average emission values from respective oil and gas facilities.
15. A system, comprising:
one or more memory modules;
one or more hardware processors communicably coupled to the one or more memory modules, the one or more hardware processors configured to execute instructions stored on the one or more memory models to perform operations comprising:
generating at least one 3D model corresponding to a real-world facility and at least one 3D model corresponding to gas emission plumes;
building a virtual environment comprising a 3D field based on the at least one 3D model corresponding to the real-world facility and the at least one 3D model corresponding to gas emission plumes;
capturing images or video of the 3D field via simulation of the virtual environment;
assigning labels to objects in the images or videos to generate labeled images or videos; and
training machine learning models using training samples including the labeled images or videos.
16. The system of claim 15 , wherein simulation of the virtual environment comprises simulation of the 3D field via drone based image capture at a predetermined angle of image capture.
17. The system of claim 15 , wherein simulation of the virtual environment comprises simulation of the 3D field via drone based video capture along as the drone traverses a predetermined path.
18. The system of claim 15 , wherein the images or video of the 3D field comprise varying spectral bands.