Files
momentry_core/scripts/ocr_processor.py

166 lines
4.6 KiB
Python
Executable File

#!/opt/homebrew/bin/python3.11
"""
OCR Processor - Text Recognition
Uses EasyOCR (local model)
"""
import sys
import json
import argparse
import os
import signal
sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
from redis_publisher import RedisPublisher
def signal_handler(signum, frame):
print(f"OCR: Received signal {signum}, exiting...")
sys.exit(1)
def process_ocr(video_path: str, output_path: str, uuid: str = ""):
"""Process video for OCR using EasyOCR"""
# Set up signal handlers
signal.signal(signal.SIGTERM, signal_handler)
signal.signal(signal.SIGINT, signal_handler)
publisher = RedisPublisher(uuid) if uuid else None
if publisher:
publisher.info("ocr", "OCR_START")
try:
import easyocr
except ImportError:
if publisher:
publisher.error("ocr", "easyocr not installed")
result = {"frame_count": 0, "fps": 0.0, "frames": []}
if publisher:
publisher.complete("ocr", "0 frames")
with open(output_path, "w") as f:
json.dump(result, f, indent=2)
return result
if publisher:
publisher.info("ocr", "OCR_LOADING_MODEL")
# Load EasyOCR reader
# languages: add more like 'fr', 'de', 'ja', 'ko', etc.
# gpu: set to True if GPU available
reader = easyocr.Reader(["en"], gpu=False, verbose=False)
if publisher:
publisher.info("ocr", "OCR_MODEL_LOADED")
# Get video info
import cv2
cap = cv2.VideoCapture(video_path)
fps = cap.get(cv2.CAP_PROP_FPS)
total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
cap.release()
if publisher:
publisher.info("ocr", f"fps={fps}, frames={total_frames}")
publisher.progress("ocr", 0, total_frames, "Starting")
# Process every N frames to speed up
sample_interval = 30 # Process every 30 frames
frames = []
frame_count = 0
processed = 0
cap = cv2.VideoCapture(video_path)
while True:
ret, frame = cap.read()
if not ret:
break
frame_count += 1
# Sample frames
if frame_count % sample_interval != 0:
continue
processed += 1
timestamp = (frame_count - 1) / fps if fps > 0 else 0
# Convert BGR to RGB
frame_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
# Run OCR
try:
detections = reader.readtext(
frame_rgb, text_threshold=0.5, low_text=0.3, link_threshold=0.3
)
except Exception as e:
if publisher:
publisher.error("ocr", f"Frame {frame_count}: {e}")
detections = []
texts = []
for detection in detections:
det: tuple = tuple(detection)
bbox = list(det[0])
text: str = str(det[1])
confidence: float = float(det[2])
x = int(min(float(p[0]) for p in bbox))
y = int(min(float(p[1]) for p in bbox))
width = int(max(float(p[0]) for p in bbox) - x)
height = int(max(float(p[1]) for p in bbox) - y)
if text.strip():
texts.append(
{
"text": text,
"x": x,
"y": y,
"width": width,
"height": height,
"confidence": confidence,
}
)
# Only add frames with text
if texts:
frames.append(
{
"frame": frame_count - 1,
"timestamp": round(timestamp, 3),
"texts": texts,
}
)
if publisher:
publisher.progress(
"ocr",
processed,
total_frames // sample_interval,
f"Frame {frame_count}",
)
cap.release()
result = {"frame_count": total_frames, "fps": fps, "frames": frames}
with open(output_path, "w") as f:
json.dump(result, f, indent=2)
if publisher:
publisher.complete("ocr", f"{len(frames)} frames with text")
return result
if __name__ == "__main__":
parser = argparse.ArgumentParser(description="OCR Text Recognition")
parser.add_argument("video_path", help="Path to video file")
parser.add_argument("output_path", help="Output JSON path")
parser.add_argument("--uuid", "-u", help="UUID for Redis progress", default="")
args = parser.parse_args()
process_ocr(args.video_path, args.output_path, args.uuid)