Tru_Image_Classifier / trufor_runner.py
Jatin-tec
crop footer
4c4f437
from __future__ import annotations
import logging
import os
import shutil
import subprocess
import tempfile
from dataclasses import dataclass
from pathlib import Path
from typing import Dict, Optional, Tuple
import numpy as np
from PIL import Image
LOGGER = logging.getLogger(__name__)
class TruForUnavailableError(RuntimeError):
"""Raised when the TruFor assets are missing or inference fails."""
@dataclass
class TruForResult:
score: Optional[float]
map_overlay: Optional[Image.Image]
class TruForEngine:
"""Wrapper that executes TruFor inference through docker or python backends."""
def __init__(
self,
repo_root: Optional[Path] = None,
weights_path: Optional[Path] = None,
device: str = "cpu",
) -> None:
self.base_dir = Path(__file__).resolve().parent
self.device = device
self.backend: Optional[str] = None
self.status_message = "TruFor backend not initialized."
backend_pref = os.environ.get("TRUFOR_BACKEND", "auto").lower()
if backend_pref not in {"auto", "native", "docker"}:
backend_pref = "auto"
errors: list[str] = []
if backend_pref in {"auto", "native"}:
try:
self._configure_native_backend(repo_root, weights_path)
self.backend = "native"
self.status_message = "TruFor ready (bundled python backend)."
except TruForUnavailableError as exc:
errors.append(f"Native backend unavailable: {exc}")
if backend_pref == "native":
raise
if self.backend is None and backend_pref in {"auto", "docker"}:
try:
self._configure_docker_backend()
self.backend = "docker"
self.status_message = f'TruFor ready (docker image "{self.docker_image}").'
except TruForUnavailableError as exc:
errors.append(f"Docker backend unavailable: {exc}")
if backend_pref == "docker":
raise
if self.backend is None:
raise TruForUnavailableError(" | ".join(errors) if errors else "TruFor backend unavailable.")
# ------------------------------------------------------------------
# Backend configuration helpers
# ------------------------------------------------------------------
def _configure_docker_backend(self) -> None:
if shutil.which("docker") is None:
raise TruForUnavailableError("docker CLI not found on PATH.")
test_docker_dir = self.base_dir / "test_docker"
if not test_docker_dir.exists():
raise TruForUnavailableError("test_docker directory not found in workspace.")
image_name = os.environ.get("TRUFOR_DOCKER_IMAGE", "trufor")
inspect = subprocess.run(
["docker", "image", "inspect", image_name],
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
check=False,
)
if inspect.returncode != 0:
raise TruForUnavailableError(
f'Docker image "{image_name}" not found. Build it with "bash test_docker/docker_build.sh".'
)
weights_candidate = Path(os.environ.get("TRUFOR_DOCKER_WEIGHTS", self.base_dir / "weights")).expanduser()
weight_file = weights_candidate / "trufor.pth.tar"
self.docker_weights_dir: Optional[Path]
self.docker_weights_dir = weight_file.parent if weight_file.exists() else None
self.docker_runtime = os.environ.get("TRUFOR_DOCKER_RUNTIME")
gpu_pref = os.environ.get("TRUFOR_DOCKER_GPU")
if gpu_pref is None:
gpu_pref = "-1" if self.device == "cpu" else "0"
self.docker_gpu = gpu_pref
gpus_arg = os.environ.get("TRUFOR_DOCKER_GPUS_ARG")
if not gpus_arg and gpu_pref not in {"-1", "cpu", "none"}:
gpus_arg = "all"
self.docker_gpus_arg = gpus_arg
self.docker_image = image_name
def _configure_native_backend(self, _repo_root: Optional[Path], weights_path: Optional[Path]) -> None:
try:
from trufor_native import TruForBundledModel
except ImportError as exc: # pragma: no cover - packaging guard
raise TruForUnavailableError("Bundled TruFor modules are not available.") from exc
default_weights = self.base_dir / "weights" / "trufor.pth.tar"
weight_candidate = weights_path or os.environ.get("TRUFOR_WEIGHTS") or default_weights
weight_path = Path(weight_candidate).expanduser()
if not weight_path.exists():
raise TruForUnavailableError(
f"TruFor weights missing at {weight_path}. Place trufor.pth.tar under weights/ or set TRUFOR_WEIGHTS."
)
try:
self.native_model = TruForBundledModel(weight_path, device=self.device)
except Exception as exc: # pragma: no cover - propagate detailed failure
raise TruForUnavailableError(f"Failed to initialise bundled TruFor model: {exc}") from exc
# ------------------------------------------------------------------
# Public API
# ------------------------------------------------------------------
def infer(self, image: Image.Image) -> TruForResult:
if image is None:
raise TruForUnavailableError("No image supplied to TruFor inference.")
prepared_image, cropped = self._strip_gps_overlay(image)
if cropped:
LOGGER.debug(
"Cropping %d px GPS overlay before TruFor inference.",
image.height - prepared_image.height,
)
if self.backend == "docker":
return self._infer_docker(prepared_image)
if self.backend == "native":
return self._infer_native(prepared_image)
raise TruForUnavailableError("TruFor backend not configured.")
# ------------------------------------------------------------------
# Inference helpers
# ------------------------------------------------------------------
def _infer_native(self, image: Image.Image) -> TruForResult:
outputs = self.native_model.predict(image)
map_overlay = None
try:
map_overlay = self._apply_heatmap(image, outputs.tamper_map)
except Exception as exc: # pragma: no cover - visualisation fallback
LOGGER.debug("Failed to build tamper heatmap: %s", exc)
return TruForResult(
score=outputs.detection_score,
map_overlay=map_overlay,
)
def _infer_docker(self, image: Image.Image) -> TruForResult:
with tempfile.TemporaryDirectory(prefix="trufor_docker_") as workdir:
workdir_path = Path(workdir)
input_dir = workdir_path / "data"
output_dir = workdir_path / "data_out"
input_dir.mkdir(parents=True, exist_ok=True)
output_dir.mkdir(parents=True, exist_ok=True)
input_path = input_dir / "input.png"
image.convert("RGB").save(input_path)
cmd = ["docker", "run", "--rm"]
if self.docker_runtime:
cmd.extend(["--runtime", self.docker_runtime])
gpu_flag = str(self.docker_gpu)
if gpu_flag.lower() in {"cpu", "none"}:
gpu_flag = "-1"
if gpu_flag != "-1" and self.docker_gpus_arg:
cmd.extend(["--gpus", self.docker_gpus_arg])
cmd.extend([
"-v",
f"{input_dir.resolve()}:/data:ro",
"-v",
f"{output_dir.resolve()}:/data_out:rw",
])
if self.docker_weights_dir is not None:
cmd.extend([
"-v",
f"{self.docker_weights_dir.resolve()}:/weights:ro",
])
cmd.append(self.docker_image)
cmd.extend(
[
"-gpu",
gpu_flag,
"-in",
"data/input.png",
"-out",
"data_out",
]
)
LOGGER.debug("Running TruFor docker command: %s", " ".join(cmd))
result = subprocess.run(
cmd,
text=True,
capture_output=True,
check=False,
)
return self._process_results(result, output_dir, image)
# ------------------------------------------------------------------
# Result parsing
# ------------------------------------------------------------------
def _process_results(self, run_result: subprocess.CompletedProcess[str], output_dir: Path, image: Image.Image) -> TruForResult:
if run_result.returncode != 0:
stderr_tail = "\n".join(run_result.stderr.strip().splitlines()[-8:]) if run_result.stderr else ""
LOGGER.error("TruFor stderr: %s", stderr_tail)
raise TruForUnavailableError(
"TruFor inference failed. Inspect dependencies and stderr:\n" + stderr_tail
)
npz_files = list(output_dir.rglob("*.npz"))
if not npz_files:
stdout_tail = "\n".join(run_result.stdout.strip().splitlines()[-8:]) if run_result.stdout else ""
raise TruForUnavailableError(
"TruFor inference produced no output files. Stdout tail:\n" + stdout_tail
)
data = np.load(npz_files[0], allow_pickle=False)
tamper_map = data.get("map")
score = float(data["score"]) if "score" in data.files else None
map_overlay = None
try:
map_overlay = self._apply_heatmap(image, tamper_map) if tamper_map is not None else None
except Exception as exc: # pragma: no cover
LOGGER.debug("Failed to build tamper heatmap: %s", exc)
return TruForResult(
score=score,
map_overlay=map_overlay,
)
@staticmethod
def _apply_heatmap(base: Image.Image, data: np.ndarray, alpha: float = 0.55) -> Image.Image:
base_rgb = base.convert("RGB")
if data is None or data.ndim != 2:
raise ValueError("Expected a 2D map from TruFor")
data = np.asarray(data, dtype=np.float32)
if np.allclose(data.max(), data.min()):
norm = np.zeros_like(data, dtype=np.float32)
else:
norm = (data - data.min()) / (data.max() - data.min())
heat = np.zeros((*norm.shape, 3), dtype=np.uint8)
heat[..., 0] = np.clip(norm * 255, 0, 255).astype(np.uint8)
heat[..., 1] = np.clip(np.sqrt(norm) * 255, 0, 255).astype(np.uint8)
heat[..., 2] = np.clip((1.0 - norm) * 255, 0, 255).astype(np.uint8)
heat_img = Image.fromarray(heat, mode="RGB").resize(base_rgb.size, Image.BILINEAR)
return Image.blend(base_rgb, heat_img, alpha)
@staticmethod
def _strip_gps_overlay(image: Image.Image) -> Tuple[Image.Image, bool]:
gray = np.asarray(image.convert("L"), dtype=np.uint8)
hsv = np.asarray(image.convert("HSV"), dtype=np.uint8)
hue = hsv[..., 0] / 255.0
sat = hsv[..., 1] / 255.0
val = hsv[..., 2] / 255.0
height, width = gray.shape
min_overlay = max(int(height * 0.08), 40)
max_overlay = max(int(height * 0.45), min_overlay + 1)
if height <= min_overlay:
return image, False
start_row = height - min_overlay
stop_row = max(height - max_overlay, 1)
row_means = gray.mean(axis=1)
sat_means = sat.mean(axis=1)
blue_mask = (hue >= 0.5) & (hue <= 0.75) & (sat >= 0.25) & (val <= 0.95)
yellow_mask = (hue >= 0.08) & (hue <= 0.18) & (sat >= 0.35) & (val >= 0.45)
white_mask = (val >= 0.87) & (sat <= 0.28)
dark_mask = val <= 0.28
overlay_mask = blue_mask | yellow_mask | white_mask | dark_mask
overlay_ratio_rows = overlay_mask.mean(axis=1)
boundary = None
best_score = 0.0
# First, try to detect a long contiguous overlay band using hysteresis on coverage.
high_ratio = 0.52
low_ratio = 0.36
run_len = 0
run_top = height
for row in range(height - 1, stop_row - 1, -1):
ratio = overlay_ratio_rows[row]
if ratio >= high_ratio or (run_len > 0 and ratio >= low_ratio):
run_len += 1
run_top = row
if run_len >= max_overlay:
break
elif run_len > 0:
if run_len >= min_overlay:
break
run_len = 0
run_top = height
elif height - row >= max_overlay:
break
if run_len >= min_overlay:
boundary_candidate = run_top
overlay_consistency = overlay_ratio_rows[boundary_candidate:height].mean()
boundary_strength = abs(row_means[boundary_candidate - 1] - row_means[boundary_candidate]) if boundary_candidate > 0 else abs(row_means[boundary_candidate] - row_means[min(boundary_candidate + 1, height - 1)])
if overlay_consistency >= 0.45 and boundary_strength >= 4.0:
overlay_height = height - boundary_candidate
margin = min(max(int(overlay_height * 0.25), 18), boundary_candidate)
crop_row = max(boundary_candidate - margin, 0)
cropped_image = image.crop((0, 0, width, crop_row))
return cropped_image, True
# Detect GPS Map Camera overlay at the bottom and crop it out if present.
for row in range(start_row, stop_row - 1, -1):
overlay_height = height - row
if overlay_height < min_overlay:
continue
if overlay_height > max_overlay:
break
overlay_hue = hue[row:height, :]
overlay_sat = sat[row:height, :]
overlay_val = val[row:height, :]
high_sat_ratio = float((overlay_sat > 0.3).mean())
dark_ratio = float((overlay_val < 0.3).mean())
bright_ratio = float((overlay_val > 0.88).mean())
colored_band_ratio = float(((overlay_sat > 0.32) & (overlay_val > 0.25) & (overlay_val < 0.85)).mean())
blue_ratio = float(((overlay_hue > 0.48) & (overlay_hue < 0.74) & (overlay_sat > 0.28)).mean())
yellow_ratio = float(((overlay_hue > 0.07) & (overlay_hue < 0.2) & (overlay_sat > 0.35) & (overlay_val > 0.45)).mean())
prev_mean = row_means[row - 1] if row > 0 else row_means[row]
boundary_strength = abs(prev_mean - row_means[row])
saturation_jump = sat_means[row - 1] - sat_means[row] if row > 0 else 0.0
score = 0.0
if high_sat_ratio > 0.38:
score += 0.8
if colored_band_ratio > 0.35:
score += 0.7
if blue_ratio > 0.22:
score += 0.8
if yellow_ratio > 0.12:
score += 0.5
if dark_ratio > 0.32:
score += 0.6
if bright_ratio > 0.07:
score += 0.5
if boundary_strength > 5.5:
score += 0.6
if saturation_jump < -0.05:
score += 0.4
edge_ratio = boundary_strength / max(row_means[row], 1.0)
if edge_ratio > 0.11:
score += 0.3
if overlay_ratio_rows[row:height].mean() > 0.42:
score += 0.3
if score > best_score:
best_score = score
boundary = row
if boundary is None or best_score < 1.6:
return image, False
overlay_height = height - boundary
margin = min(max(int(overlay_height * 0.25), 18), boundary)
crop_row = max(boundary - margin, 0)
cropped_image = image.crop((0, 0, width, crop_row))
return cropped_image, True