We have an image database and we want to do object recognition algorithm by neural network. For this we need to divide the image to some path and then extract patch to feature extraction. What is the best size of patch?
The size of patches strongly depend on the scene and type of object. it's better to use multiresolution methods and choose best size according to the result of this process.
The size of patches strongly depend on the scene and type of object. it's better to use multiresolution methods and choose best size according to the result of this process.