Hello, good evening all,
I’m doing an Object based image classification, i used LSMS segmentation to do the segmentation, I also created 5 training samples but Vector Classifier returned just one class.
What could be the problem
Dear @Treasurepeace1,
Could you share with us the logs of the processing? And also tell us which version of OTB you are using and how you run the applications (command line, malpa, QGIS)?
I used OTB 7.0.0 standalone
The first image is the log for the Vector classifier and the second is the train-vector classifier
It appears that the result of the training step was not successful. Indeed, according to the logs, all the indicators (precision, recall, F-score) a set to zero. Are you sure your training set is not empty? May share a screen shot of the input to give to the trainer?
I’m attempting to repeat the process and I’m having this error. This us for the train vector classifier
I attempted to repeat the process and I got this error. This is for the train vector classifier…
The TrainVectorClassifier takes a vector dataset as input. Providing a tif file doesn’t seems like a good idea. We usually use shp, geojson, gpkg or sqlite files as input.
Thanks @julienosman I’ve corrected it.
But I had this error while using the vector classifier
This is the input data
This the log from the input data
Log from monteverdi application