feat: Improve documentation for unprojection functions in camera module

- Enhanced docstrings for `unproject_points_to_z_plane` and the corresponding method in the `Camera` class to provide detailed descriptions of arguments and return values.
- Clarified the purpose and usage of the unprojection functionalities, improving overall code readability and usability.

app/camera/__init__.py

@@ -155,6 +155,19 @@ def unproject_points_to_z_plane(
     pose_matrix: Float[Array, "4 4"],
     z: float = 0.0,
 ) -> Float[Array, "N 3"]:
+    """
+    Un-project 2D points to 3D points on a plane at z = constant.
+
+    Args:
+        points_2d: 2D points in pixel coordinates
+        K: Camera intrinsic matrix
+        dist_coeffs: Distortion coefficients
+        pose_matrix: Camera-to-World (C2W) transformation matrix
+        z: z-coordinate of the plane (default: 0.0,  i.e. ground/horizon/floor plane)
+
+    Returns:
+        [..., 3] world-space intersection points
+    """
     plane_normal = jnp.array([0.0, 0.0, 1.0])
     plane_point = jnp.array([0.0, 0.0, z])
     return unproject_points_onto_plane(
@@ -374,7 +387,14 @@ class Camera:
         self, points_2d: Num[Array, "N 2"], z: float = 0.0
     ) -> Num[Array, "N 3"]:
         """
-        Unproject 2D points to 3D points on a plane at z = constant.
+        Un-project 2D points to 3D points on a plane at z = constant.
+
+        Args:
+            points_2d: 2D points in pixel coordinates
+            z: z-coordinate of the plane (default: 0.0,  i.e. ground/horizon/floor plane)
+
+        Returns:
+            [..., 3] world-space intersection points
         """
         return unproject_points_to_z_plane(
             points_2d,