I’m a newbie in remote sensing, QGIS and OTB and currently I’m working on Object based image classification using a raster images captured by UAV with Micasense RedEdge-MX. I’m trying to classify images from agricultural areas and I’m trying to follow the methods presented in this video (Thanks to Mr. Jesse!) Object-based Image classification in QGIS || OBIA !! || A complete Tutorial - YouTube.
So far I have no problem from the start except in “Vector classifier”. I have generated all the info and layers prior to “Vector classifier” (e.g. classes, training samples, segmentation result, segmentation statistics using zonal stat, joined layer (by location), and the models generated from train vector classifier. When I try the Vector classifier using segmentation stat.shp and model as inputs, and field names to be calculated, this error occurs: " (FATAL) VectorClassifier: itk::ERROR: Failed to open GDALDataset file C:\Users"
I am using OTB vs 7.3 with OTB master plugin in QGIS 3.18.1 Zurich (Windows 10 64-bit) which I installed few days ago.
I hope someone can help me.
Also, do you have suggestions of other methods for image classification of UAV images?
When you use QGIS to launch an OTB processing, you need to avoid spaces in path to files. Can you check that there is no space in the path to the vector data? If so, you can to try to add quotes " " around the path.
Sorry I forgot to reply back immediately. But yes, I was able to execute the vector classification. Thanks to your instruction and additional information.
The problem relates to the presence of a space in the User folder “EJEH CHISOM”, There are known problems with windows handling spaces in paths. Can you try writing the classified.shp in another path where you are sure there are no spaces?
Please I need assistance, see snipped error message below