apds/utils.py

import cv2
import base64
import csv
import numpy as np
from pathlib import Path
from typing import List, Dict, Any

def draw_boxes(image: np.ndarray, detections: List[Dict[str, Any]]) -> np.ndarray:
    """
    Draws bounding boxes and confidence labels on an image array based on detections.
    
    Inputs:
        image: A numpy array representing the image (BGR format).
        detections: A list of dictionaries, each containing 'xyxy' (list of [x1, y1, x2, y2]), 
                    'confidence' (float), and 'class_id' (int).
    
    Outputs:
        A copy of the input numpy array with bounding boxes plotted.
        
    Side effects:
        Does not mutate the original image array; returns a new annotated copy.
    """
    img_copy = image.copy()
    for det in detections:
        # Convert bounding box coordinates to integers since cv2 expects integer pixels
        x1, y1, x2, y2 = map(int, det['xyxy'])
        conf = det['confidence']
        
        # Pothole label construction
        label = f"Pothole: {conf:.2f}"
        
        # Draw the rectangle
        cv2.rectangle(img_copy, (x1, y1), (x2, y2), (0, 0, 255), 2)
        # Place the text above the bounding box
        cv2.putText(img_copy, label, (x1, max(y1 - 10, 10)), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 2)
        
    return img_copy

def image_to_base64(image: np.ndarray) -> str:
    """
    Converts a numpy image array to a base64 encoded string.
    
    Inputs:
        image: A numpy array representing the image.
        
    Outputs:
        A base64 encoded string of the compressed PNG version of the image.
        
    Side effects:
        None.
    """
    success, buffer = cv2.imencode('.png', image)
    if not success:
        return ""
    b64_str = base64.b64encode(buffer).decode('utf-8')
    return b64_str

def export_csv(detections: List[Dict[str, Any]], export_path: str | Path) -> None:
    """
    Writes a list of detection dictionaries to a CSV file.
    
    Inputs:
        detections: A list of detection dictionaries containing flattened key-value pairs 
                    (e.g., filename, confidence, latitude, etc.).
        export_path: The file path (string or pathlib.Path) where the CSV will be saved.
        
    Outputs:
        None.
        
    Side effects:
        Creates or overwrites a file at the specified absolute export_path. 
        Creates parent directories iteratively if they do not exist.
    """
    file_path = Path(export_path)
    # Ensure immediate directory exists
    file_path.parent.mkdir(parents=True, exist_ok=True)
    
    if not detections:
        file_path.touch()
        return

    # Extract dynamic keys efficiently
    keys = list(detections[0].keys())
    
    with file_path.open('w', newline='', encoding='utf-8') as f:
        dict_writer = csv.DictWriter(f, fieldnames=keys)
        dict_writer.writeheader()
        dict_writer.writerows(detections)