Annie’s Journey with free AI | Annie 與 免費AI 的日常

import mediapipe as mp
from mediapipe.tasks import python
from mediapipe.tasks.python import vision
# ==========================================
# Annie's AI Lab - Object Detection Script
# Model: EfficientDet-Lite0
# ==========================================
def run_detection():
    # 1. Configuration: Set the model path
    # Ensure 'efficientdet_lite0.tflite' is in the same directory
    model_path = 'efficientdet_lite0.tflite'
    # 2. Detector Options Setup
    # score_threshold: Lowering this might help find the "hidden" celery!
    base_options = python.BaseOptions(model_asset_path=model_path)
    options = vision.ObjectDetectorOptions(
        base_options=base_options,
        score_threshold=0.3, 
        running_mode=vision.RunningMode.IMAGE
    )
    # 3. Create Detector Instance
    with vision.ObjectDetector.create_from_options(options) as detector:
        
        # 4. Load the input image
        # Ensure the filename matches your photo (e.g., 'test.jpeg')
        image_file = 'test.jpeg' 
        try:
            numpy_image = mp.Image.create_from_file(image_file)
        except Exception as e:
            print(f"Error: Could not load image file '{image_file}'. {e}")
            return
        # 5. Execute Inference
        detection_result = detector.detect(numpy_image)
        # 6. Output Analysis Report
        print("\n" + "="*45)
        print(" ANNIE'S PYTHON AI LAB - SCAN REPORT 🐰✨")
        print("="*45)
        
        if not detection_result.detections:
            print("Status: No objects detected above threshold.")
            print("Action: Try lowering the score_threshold in the code.")
        else:
            print(f"Status: {len(detection_result.detections)} object(s) found.")
            print("-" * 45)
            
            for i, detection in enumerate(detection_result.detections):
                category = detection.categories[0]
                label = category.category_name.upper()
                confidence = round(category.score * 100, 2)
                
                # Retrieve Bounding Box Coordinates
                bbox = detection.bounding_box
                
                print(f"[{i+1}] Object: {label}")
                print(f"    Confidence: {confidence}%")
                print(f"    Location: [X:{bbox.origin_x}, Y:{bbox.origin_y}, W:{bbox.width}, H:{bbox.height}]")
                print("-" * 45)
        print("UAT Result: Scan completed successfully.")
if __name__ == "__main__":
    run_detection()