I was using blender to simulate a structured light setup consisting of a camera and a projector. To reconstruct an object/scene, I need to know the focal length of the camera and the projector. I figured out the camera focal length, but I’m having a hard time figuring out the projector focal length. Can someone give me a hint? thanks.
I’m only used to dealing with throw ratio, which is basically image width / throw distance, and it’s range is typically dictated by the lens. I have to admit, I’ve never setup an actual projection, I’ll do a volumetric light pyramid and simply map the screen without bothering about accurate shadows (since we tend to try really hard to avoid those :D). But I try to model response to ambient lighting (screen gain, coating if any etc) and light attenuation (bigger image means less bright using inverse square law), but ignore attenuation by color wheel, blending (if any), and lens distortion.
If I were to try this, I’d setup my lens and image plane. Position 3D cursor on lens. Copy the image plane and scale it about the 3D cursor until it falls on the lens. That’s the basis to create the light pyramid (rectangle on the screen and rectangle on the lens). Duplicate the edges (to separate object) connecting the two and extend them to where they meet. This is where your size 0 light point will be - somewhere in the lens but not at the chip - we don’t want to calculate light paths done by the lens. Not possible without some lens design software (not cheap).
For accuracy you want to setup a real test; read out the throw ratio if you can, and measure the size of the image on the screen and in the lens.
I did figure out a way to stack several images on top of each other on the lens with different intensities, size, and offsets, to indicate many levels of glass in there, but I can’t find the file.
You have to share more about what you’re actually trying to achieve here.
What I want to do is simulate the structured light system. I use a spot light source as the projector and assign an image texture to the light source. It works well as a projector, but I just couldn’t figure out how to compute the focal length of this projector
I have an idea about the parameters. the elements in matrices of internal parameters can be seen as a number related to the ratio of focal length and pixel length。You can regard it as tan(f/alpha). By the way, the projected pattern can not keep it‘s original aspect ratio.I’m simulating the structured light theory and seeking for the solution of projecting under parameterization control. Do you have any idea about that?