Most cameras have some sensitivity to IR as well as to visible light. Color cameras used colored filters which block IR, and monochrome (black and white) cameras may have IR blocking filters, but these filters may be removable (e.g., a color camera with a monochrome "night mode" will simply move the colored filter out of the way so its sensor will detect everything). A camera designed for IR may have better sensitivity to IR, but the basic functionality is the same.
A common misconception is that IR cameras detect heat. Hot objects do emit IR, but at temperatures you will encounter in a behavior lab, that is mid to far infrared (wavelengths of 3000 nm and up) which is not detectable by your camera (even IR cameras typically have sensitivity only up to around 1000 nm). Typical IR illumination is 850 or 940 nm (e.g., 940 nm in the Phenotyper), in the “near IR” range (defined as 700 nm to 3000 nm). To be visible in the near IR range simply based on heat, an object would have to be over 1800 degrees C. It’s possible to detect the heat from living things, but for that you need a thermographic camera and not merely a camera that is sensitive to near IR. Thermographic cameras add expense and typically have low resolution, resulting in a less detailed image, that shows temperature as false color; that's not something normally used for behavior tracking.
So, for purposes of behavior research, IR is essentially just a color of light that has the special trait that it is invisible to us and to most animals (including rodents). We add IR illuminators that work just like visible lights, but in this invisible color. Those illuminators may be separate devices, mounted in an observation cubicle, or built into the top of our Phenotyper chambers. In some cases we will add an IR-pass filter which prevents the camera from seeing visible light; only IR is able to pass through the filter and be detected. This permits the camera to give a consistent image even if the ambient lighting is changing (e.g., circadian lighting), but requires that a source of IR illumination is present. You can see that this filter is present because the front of the camera lens will be completely black, as no visible light passes through it.
When using IR illumination, all the same considerations as using normal visible light apply: light is shining onto the scene, it is reflected or absorbed by objects in the scene, and whatever is reflected back to the camera becomes part of the image. If there is a shiny surface, IR can produce bright glare, and the relative position of the camera and illuminator may need to be adjusted to get the best possible image. For the most part, dark objects will appear dark and light objects will appear light in the camera image, but there are some exceptions as some objects will absorb more or less light in the IR range than they do in the visible range (just like some objects will absorb more or less red light, and accordingly may appear blue or red).
If you are using IR illumination to observe animals on bedding, corncob bedding may appear a bit brighter in IR than it does to the eye, and we find that paper-based bedding often appears darker (although not all brands will appear dark). For light-colored animals it can be advantageous to use very dark bedding. One option is Purina’s Yesterday’s News kitty litter (unscented!) which appears very dark.