About NeoFract
NeoFract (Neuron Fracture) is an advanced software solution that utilizes Machine Learning algorithms, specifically Convolutional Neural Networks (CNN), to analyze and estimate fracture properties from images with unprecedented accuracy.
Fractures are common in nature, appearing as cracks in soil, rocks, and subsurface fluid flow paths. While they may appear as simple lines at first glance, fracture geometries are actually complex with variations influenced by parameters like surface roughness and mean aperture (gap width).
Traditional methods of fracture analysis can be time-consuming and subjective. NeoFract provides a rapid, objective solution for estimating these critical fracture parameters with high accuracy MAPE <10%.
How It Works
NeoFract employs a trained Convolutional Neural Network that processes fracture images to estimate:
- Surface roughness coefficient (a measure of fracture wall irregularity)
- Mean aperture (average width of the fracture gap)
Technical Approach
The CNN model analyzes the visual patterns in fracture images, learning to correlate specific image features with known physical measurements. This approach offers several advantages:
- Non-destructive analysis (works from images alone)
- Rapid processing (results in seconds)
- Consistent, objective measurements
- Ability to handle complex fracture patterns
Fracture Analysis
Upload a fracture image to estimate its properties using our Machine Learning model.
Processing image with Machine Learning model...
Analysis Results
Surface Roughness Coefficient
The surface roughness coefficient indicates the irregularity of the fracture walls. Higher values (closer to 1) suggest more complex, irregular surfaces, while lower values indicate smoother fracture walls.
Mean Aperture
The mean aperture represents the average width of the fracture opening. This measurement is crucial for understanding fluid flow potential through the fracture system.