>>58014259
First, it is very likely it will not be clear for you until you try it for yourself. You can ask me questions here.

Controlnet is a system that allows to take an image, use a preprocessor to extract information from that image, and then a model to use that information to “guide” the generation of the final image. This is made on top of the text prompt that models receive.

In the example I show, I use the preprocessor “depth_anything_v2” to generate a depth map. And the using the model “noobaiXLControlnet_epsDepthMidasV11” to influence the generation of the model “waiNSFWIllustrious”, and creating an image with similar positions and pose.

I also attach the in the image the interface in Forge.
I know that controlnet is implemented in modern UIs, is integrated by default in Forge and ReForge, need a complement in Automatic1111. Comfy also uses it, but don't know how.
You will also need the models, for Illustrious these are my favorites:
https://civitai.com/models/962537/noobai-xl-controlnet-openpose
https://civitai.com/models/929685?modelVersionId=1042508
https://civitai.com/models/929685?modelVersionId=1040650

There is this very in depth article about controlnet, but is old, so uses SD1.5 instead of Illustrious, and do not mention how is used in modern UIs, but still pretty good:
https://stable-diffusion-art.com/controlnet/#Image_Prompt_adapter_IP-adapter