Hi @user1234,
Did you follow all the installation steps of that node? That is required because it uses the canvas npm package as a dependency for fast(er) drawing of the bounding boxes on the images.
Some other nodes (like e.g. node-red-contrib-object-detection) do it in pure javascript, so you don't need to do such extra build steps. But those use the pureimage npm package which is really slow and eats a lot of cpu. Which is not usable when you want to do a lot of image processing on hardware like e.g. a Raspberry Pi.
What is your use case? Do you need a lot of image processing?
Bart