"""
Maritime_Custom - Example Inference Script
Repository: MuayThaiLegz/Maritime_Custom
"""

from ultralytics import YOLO
import cv2

# ============================================
# 1. BASIC INFERENCE
# ============================================

# Load model
model = YOLO('MuayThaiLegz/Maritime_Custom')

# Single image
results = model.predict(
    source='maritime_image.jpg',
    conf=0.25,
    save=True
)

# ============================================
# 2. VIDEO PROCESSING
# ============================================

# Process video with streaming
results = model.predict(
    source='port_surveillance.mp4',
    conf=0.25,
    stream=True,
    save=True
)

for i, result in enumerate(results):
    print(f"Frame {i}: {len(result.boxes)} vessels detected")

# ============================================
# 3. REAL-TIME WEBCAM
# ============================================

# Live detection
results = model.predict(
    source=0,  # Webcam
    conf=0.25,
    show=True,
    stream=True
)

# ============================================
# 4. BATCH PROCESSING
# ============================================

# Process directory
results = model.predict(
    source='maritime_images/',
    conf=0.25,
    save=True,
    project='detections',
    name='batch_run'
)

# ============================================
# 5. CUSTOM POST-PROCESSING
# ============================================

# Advanced detection with custom handling
results = model.predict('image.jpg', conf=0.25)

for result in results:
    boxes = result.boxes
    img = result.orig_img
    
    for box in boxes:
        # Extract info
        x1, y1, x2, y2 = box.xyxy[0].cpu().numpy()
        conf = float(box.conf[0])
        
        # Draw custom annotations
        cv2.rectangle(img, (int(x1), int(y1)), (int(x2), int(y2)), (0, 255, 0), 2)
        cv2.putText(img, f'Vessel: {conf:.2%}', (int(x1), int(y1)-10),
                    cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)
    
    cv2.imwrite('annotated.jpg', img)

# ============================================
# 6. EXPORT FOR PRODUCTION
# ============================================

# Export to ONNX
model.export(format='onnx', dynamic=True)

# Export to TensorRT (NVIDIA)
model.export(format='engine', device=0, half=True)

print("✅ Inference examples complete!")