Getting all objects being shown?

Scripting Issues
1

Sorry to bother, but is there a way to get all the stuff being rendered on the scene via the camera?

As in, a Panda3D method of sorts.

‘cuz I want to find a way to be able to detect all the nodepaths that are being shown, so that I can highlight a particular few of interest… And if theres no built-in method, I gonna have to do a check between the objects’ positions and the camera’s position, and the angle of the camera to the object… That… should work as well, right?

2

guess you just need this line:

put it before the run() command

3

Thanks for replying, but that… wasn’t exactly what I was aiming for…

I was talking about stuff which is being seen on the camera, not those that are on the scene, ehehe…

Basically, what you are looking at. So for example, I look at an apple in front of me, and I look at the wall behind me, and the results which come out will be different, one having the apple and one without.

Is it possible?

4

ah sorry not getting what you said - a possible solution is with collision solids: you provide a collision solid shooting off the camera lens and then test whos inside - as much as fishing with a net - here you’ll find all you need
PS: of course is not what you need ready to go - you gotta take off the parts you need and set the polygon boundaries for the entire frustum

5

Here’s one way to do it. In this one, you have to query each node/object- that should not be hard just set up a for loop using this function and it will return True if it’s on screen and False if it’s not.

def _onscreen(object):
	# "object" is the nodepath you are asking about. 
	# Convert the point to the 3-d space of the camera 
	p3 = base.cam.getRelativePoint(render, object.getPos()) 

	# Convert it through the lens to render2d coordinates 
	if not base.camLens.project(p3, Point2() ): 
		return False
	else: return True
6

Or simply call isInView for every object.

7

Ahh… Gonna give it a try then, all your suggestions.

Thanks alot everyone!

8

Thanks alot! mavasher, I used your method, works like a charm. Thanks once again for the code?

But out of curiousity, how do I do isInView? Could you provide me an example? Because it sounds really interesting…

9

base.camNode.isInView(point) is a more reliable way of doing what mavasher does with project().

10

I tried but with no luck: no object detected in view

*see 2 post down
11

The point has to be in the coordinate space of the camera, of course.

12

ok, now we got it:

# -*- coding: utf-8 -*-
from direct.showbase.DirectObject import DirectObject
import random, sys

from direct.directbase.DirectStart import *

base.cam.setPos(0,-20,0)

class World(DirectObject):
  def __init__(self):
    for i in range(20):
      ball = loader.loadModel("smiley")
      ball.reparentTo(render)
      ball.setPos(random.randint(-3,3)*2, random.randint(-3,3)*2, random.randint(-3,3)*2)
    self.accept("escape", sys.exit)
    self.accept("mouse1", self.click)

  def click(self):
    objs=base.render.findAllMatches("**/+ModelNode")
    for o in objs:
      if base.camNode.isInView(o.getPos(base.cam)):
        print "%r IN view"%o.getPos()
        o.setColor((1,0,0,1))
      else:
        print "%r NOT IN view"%o.getPos()
        o.setColor((1,1,1,1))

w = World()
run()

To be short I tested just the object center anyhow with a little more code using the getTightBounds method should be perfect.

13

Ahh… I see.

Just applied it, works perfectly as well.

Thanks alot everyone!