hello every one
what is the best program to apply machine learning / deep learning approaches on satellite images like Aster, Landsat, Sentinel
so, i prepare images for the required program format and get efficient and accurate results
then i could export the results for other RS programs with more easily?
Hi Solmaz Babakan
I suggest Python; it has several libraries related to image processing.
Matlab is efficient either. However, it is slower than python.
so thanks Mehran Shahpari
but can you explain how some more
I used python for some simple applications like estimations, but I can’t imagine how should I do that for satellite imagery
in fact, I get e.g. a sentinel image around a mine
I want to use this mine spectral characteristics to learn a model and classify the whole image regards the probability of possible orebodies in the region
here is a true color image of an example sentinel dataset and the mine extent
i will be so glad if you help me to understand
Dear Solmaz Babakan
Python is one of the most powerful languages for image processing; however, as you know, python needs one of the following libraries to be able to do image processing. So, you should download and install the library and call it in python by the "import" command.
- Scikit-image
- OpenCV
- Mahotas
- SimplelTK
- ....
Note that learning this process takes time to get the desired result.
hmm
I will try them
i try some simple models for multiple regression using keras, Scikit ... but i cant get any perfect results
as you know as mentioned worldwide DL is the best estimator
it could be founded various real-world examples
but my models do not perform as perfect as I expect
for example, loss, mse functions go to be stable but are high yet so the error of estimations is high inevitably
have you any suggestion where is my mistake?
I would recommend using U-Nets. They are great at image segmentation. This would be a way to solve your problem: Article U-Net: Convolutional Networks for Biomedical Image Segmentation
Cheers, Raoul
Hie Solmaz Babakan
Python provides an easy to use and efficient environment. You need to install its libraries such as TensorFlow, Keras in a Jupyter Notebook setup which then assist you with both code and text manipulation.
GEE can do that with tensorflow in colab. Language can be python or javascript.
- Badges
- Science topic