Detecting if a UI object is blocking a 3d object

Hey everyone!
I'm trying to write code that detects whether I see a 3d object or not. I combined various methods, like getComponent<Renderer>().isVisible and raycasting against the bounding points of the collider, but nothing seems to detect whether there's a UI object that's blocking (appearing in front of) it.
Physics.Raycast doesn't detect UI objects.
I can't run raycast with GraphicsRaycaster since I have a lot of UI objects in the scene and it'll be too inefficient.
I'm also using WebGL.

Anyone has any ideas on how to detect if a UI element is blocking a 3d object?

 

Not sure about checking if it's entirely blocked, but you could use WorldToViewportPoint to get its position on the screen, and check if a UI element is over that spot. It would only check for the exact position of the object though, and wouldn't account for an object's size.

 

Thanks a lot for the answer @RadRedPanda!
Can you point me to how to calculate if a UI object is over this viewport point? I can't use raycasting from that point since it doesn't detect UI objects

 

You can convert your UI element to a Rect object, and then check if the Rect has the point inside of it using Rect.Contains(). You would need to use WorldToScreenPoint instead though, so I incorrectly mentioned that in my last post.

 

This would not necessarily work for UI objects that are in world space (instead of screen overlay), since their Rect might not contain the point, but still block the 3d object from view :\

 

I didn't know you were rendering in World Space. Doesn't this mean that you have many different Canvases though? From reading the documentation, it would seem that the GraphicsRaycaster only casts on the Canvas object it is attached to, so would not necessarily be inefficient in colliding with too many stray UI objects.

 

I render in both World Space and Screen Overlay
Yea but that would mean I have to use the Graphics Raycaster for each canvas (basically call Raycast for every UI canvas I have, which is a lot)
The canvases are also dynamic, which means I'll have to track at every frame how many canvases exist (some are destroyed, some are created)

 

I don't know how realistic this is for your project, but you could just create one Canvas for all of the UI elements you want to be, and just set the positions of all of your UI elements within that Canvas space. I also suspect you're using World Space to do something similar to rendering UI above a parent object, and to get around that, you could update the position of the UI element in code to follow the object. This would probably be in the same place you update the UI elements to look at the camera.