Add comprehensive tests for image processing and mask functionalities

- Introduced tests for color image handling, ensuring source preservation and premultiplied alpha usage.
- Implemented tests for brush stroke operations, validating segment boundaries and brush anti-aliasing.
- Added tests for frame anti-aliasing, confirming correct rendering behavior with various alpha values and depth settings.
- Created tests for frame projection, ensuring accurate overlap calculations and half-plane clipping.
- Developed tests for image premultiplied retention, verifying behavior during loading and packing processes.
- Added undo functionality tests for mask operations, ensuring single transaction behavior during edits.

src/anyimage/server/app.py

@@ -16,7 +16,6 @@ from .jobs.debug_predictions import (
 )
 from .jobs.cutout import run as _generate_cutout_artifacts
 from .jobs.depth_plane import run as _generate_depth_plane_geometry
-from .jobs.refine_image_selection import run as _refine_image_selection
 from .jobs.remove_background import run as _remove_background
 from .jobs.upscale import run as _upscale_image
 from .model_manager import ModelManager
@@ -41,11 +40,6 @@ def download_required_models(context, parameters):
     return manager.download_missing(context, parameters["models"])
 
 
-@server.job("refine-image-selection")
-def refine_image_selection(context, parameters):
-    return _refine_image_selection(context, parameters)
-
-
 @server.job("remove-background")
 def remove_background(context, parameters):
     return _remove_background(context, parameters)

src/anyimage/server/geometry/normal_texture.py

@@ -40,15 +40,9 @@ def write_camera_normal_texture(
     if visibility.shape != normal.shape[:2]:
         visibility = resized_visibility(visibility, normal.shape[:2])
     normal_mode = str(normal_mode).upper()
-    if normal_mode == "TANGENT":
+    if normal_mode in {"TANGENT", "OBJECT"}:
         converted = normal * np.asarray((1.0, -1.0, -1.0), dtype=np.float32)
         empty_normal = (128, 128, 255)
-    elif normal_mode == "OBJECT":
-        converted = normal[..., (2, 0, 1)] * np.asarray(
-            (1.0, -1.0, -1.0),
-            dtype=np.float32,
-        )
-        empty_normal = (0, 128, 128)
     else:
         raise ValueError(f"Unsupported normal mode: {normal_mode}")
     encoded = np.clip(

src/anyimage/server/jobs/cutout.py

@@ -3,33 +3,20 @@
 from pathlib import Path
 
 from ..geometry.prediction_artifacts import generate_moge2_artifacts
-from ..selection import refine_selection
 
 
 def run(context, parameters):
     context.progress(0.05, "Preparing Cutout Selection")
     context.directory.mkdir(parents=True, exist_ok=True)
     input_path = Path(parameters["input"])
-    selection_image_path = input_path
     result = {}
-    if parameters.get("refine_selection"):
-        context.progress(0.2, "Refining Selection")
-        selection_image_path = refine_selection(
-            context,
-            input_path,
-            parameters["refine_model"],
-            parameters["device"],
-            context.directory / "selection.png",
-        )
-        result["selection_image"] = selection_image_path.name
-
     depth_mode = str(parameters.get("depth_mode", "NONE"))
     normal_mode = str(parameters.get("normal_mode", "NONE"))
     if depth_mode != "NONE" or normal_mode != "NONE":
         artifacts = generate_moge2_artifacts(
             context,
             parameters,
-            selection_image_path,
+            input_path,
             depth_mode=depth_mode,
             normal_mode=normal_mode,
         )

src/anyimage/server/jobs/refine_image_selection.py

@@ -1,24 +0,0 @@
-import time
-
-from ..selection import refine_selection
-
-
-def run(context, parameters):
-    """Refine foreground Alpha while preserving the submitted bounds canvas."""
-    started = time.perf_counter()
-    context.progress(0.05, "Preparing image Selection")
-    context.check_cancelled()
-    output_directory = context.directory
-    output_directory.mkdir(parents=True, exist_ok=True)
-    context.progress(0.2, f"Loading BEN2 for {parameters['device'].upper()}")
-    output_path = refine_selection(
-        context,
-        parameters["input"],
-        parameters["model"],
-        parameters["device"],
-        output_directory / "image.png",
-    )
-    elapsed_ms = (time.perf_counter() - started) * 1000.0
-    message = f"Image Selection refined in {elapsed_ms:.0f} ms"
-    context.progress(1.0, message)
-    return {"selection_image": output_path.name}

src/anyimage/server/selection/__init__.py

@@ -1,5 +0,0 @@
-"""AI generation for rasterized image Selections."""
-
-from .refine_selection import refine_selection
-
-__all__ = ("refine_selection",)

src/anyimage/server/selection/refine_selection.py

@@ -1,23 +0,0 @@
-from ..models.ben2 import infer_foreground
-
-
-def refine_selection(
-    context,
-    input_path,
-    model,
-    device,
-    output_path,
-):
-    """Refine foreground Alpha within one rectangular Selection context."""
-    model_manager = context.resource("model_manager")
-    foreground, _load_ms, _inference_ms = infer_foreground(
-        input_path,
-        model_manager.directory(model),
-        device,
-        model_cache=model_manager,
-        cancel_check=context.check_cancelled,
-        preserve_input_alpha=True,
-    )
-    context.check_cancelled()
-    foreground.convert("RGBA").save(output_path, format="PNG", compress_level=1)
-    return output_path