Refactor observation preprocessing to use a modular pipeline system

- Introduced `RobotPipeline` and `ObservationProcessor` for handling observation transformations.
- Updated `preprocess_observation` to maintain backward compatibility while leveraging the new pipeline.
- Added tests for the new processing components and ensured they match the original functionality.
- Removed hardcoded logic in favor of a more flexible, composable architecture.

tests/envs/test_envs.py

@@ -22,7 +22,9 @@ from gymnasium.utils.env_checker import check_env
 
 import lerobot
 from lerobot.envs.factory import make_env, make_env_config
-from lerobot.envs.utils import preprocess_observation
+
+from lerobot.processor.pipeline import RobotPipeline, TransitionIndex
+from lerobot.processor.observation_processor import ObservationProcessor
 from tests.utils import require_env
 
 OBS_TYPES = ["state", "pixels", "pixels_agent_pos"]
@@ -48,7 +50,12 @@ def test_factory(env_name):
     cfg = make_env_config(env_name)
     env = make_env(cfg, n_envs=1)
     obs, _ = env.reset()
-    obs = preprocess_observation(obs)
+    
+    # Process observation using pipeline
+    obs_pipeline = RobotPipeline([ObservationProcessor()])
+    transition = (obs, None, None, None, None, None, None)
+    processed_transition = obs_pipeline(transition)
+    obs = processed_transition[TransitionIndex.OBSERVATION]
 
     # test image keys are float32 in range [0,1]
     for key in obs: