-
Inspect the camera
- Called camera_get_pose to obtain the camera position, rotation, forward vector, and FOV.
-
Find a valid placement surface
- Calculated a point in front of and below the camera.
- Called scene_raycast_down to locate a walkable ground surface without using a settle tool.
-
Spawn the character
- Called character_spawn_ready to create AppearanceCharacter at the detected ground position.
-
Aim the camera
- Called camera_look_at to center the character in the camera view.
-
Discover the Blueprint interface
- Called blueprint_inspect_api and identified set_app(NewParam) as the character appearance function.
-
Switch and capture appearances
- Repeatedly called blueprint_call with set_app values 1–10.
- After each appearance change, called scene_capture_view with the mvrc source to save a PNG screenshot.
-
Process the outputs
-
Copied all captured images into examples/ change_character_appearance.
-
Used Python and OpenCV (cv2) to convert every image to BGRA and set the complete alpha channel to 255.
-
Verified all 11 generated images had four channels and fully opaque alpha values.
-