Hi, cool!
I am also kind of new, some things to consider if you struggle to see the images in your locations:
Connect Your stories to the locations more directly, if the location is a desk, have your image interact with it more directly, it can add an extra context to the locus. If it is a bear for example, have the bear eat and destroy the desk in your mind, instead of just putting a bear on the desk. This can sacrifice speed a bit for higher level performances, because you need to form those extra connections without encoding more information, but it has helped me a bit, at least in the very beginning.
In this example, maybe you think: oh no! The bear is destroying the desk! This can also include your emotions and will add another, non-visual layer to the encoding - a feeling attached to the scene happening at that location. This can be useful with other emotions also: something can be funny, sad, embarrassing, fascinating, dangerous, crazy, smell bad/good, etc. When you get to the locus in your mind, maybe you can remember how the image/scene made you feel when you made it? This can be enough of a trigger to remind you of what the image/scene was.
This may or may not apply to you, but when you make locations, make them distinct. At least if they are close to one another. For example, if you have a room With 3 chairs, it can be confusing to use all of them as 3 locations, you might mix up which image go where. For example, I have a building with some offices, but if there are 4 small, very similar office rooms next to each other. I might only use 2 of them, the ones furthest apart, and find something distinct in them, plants, ornaments to make each room unique. Some folks would just use all 4 offices, no problem, you just have to try and see what Works.