@Lavagnor, for the poster:
Future Work
This pilot study shows that CNN has the potential to estimate turbidity in smaller rivers using remote sensing images; something that was previously conducted on larger rivers with multispectral image band formulae. This can be expanded to sites around the world. Once a CNN is trained on a site, we will apply it to the same site in a different season and to sites in different climates (e.g., a CNN trained in semi-arid South Africa applied to the southeast United States). This could elucidate differences in how multispectral images detect turbidity or other signs of water clarity.
@Lavagnor, for the poster:
Future Work
This pilot study shows that CNN has the potential to estimate turbidity in smaller rivers using remote sensing images; something that was previously conducted on larger rivers with multispectral image band formulae. This can be expanded to sites around the world. Once a CNN is trained on a site, we will apply it to the same site in a different season and to sites in different climates (e.g., a CNN trained in semi-arid South Africa applied to the southeast United States). This could elucidate differences in how multispectral images detect turbidity or other signs of water clarity.