Determination of water turbidity using image processing techniques
- 0 Collaborators
The intention behind this project is implementing Image processing techniques like Image recognition and classification to determine the level of turbidity in water samples. The most broadly used methods involves long waiting times for results and cumbersome calibration procedures besides from the fact of using, sometimes, expensive equipment. The intended methodology would be taking pictures of water samples with a high definition camera in a controlled environment, and using to identify the level of turbidity on the taken sample a previously trained network. ...learn more
Project status: Concept
Overview / Usage
- Water analysis is an ubiquitous necessity around the world, and for many communities in developing countries, like Colombia for example, access to specialize equipment sometimes is not a possibility. Therefore, having a simpler way to measure accurately the parameter of turbidity using a smartphone camera (further stage of the project) could be very advantageous for them.
Methodology / Approach
- Development of different neural network architectures for image recognition
- Training of the neural network with images of samples from different devices with known turbidity value
- Taking of samples to be measured
- Recognition and classification of the samples images
- Evaluation of results