如何使用 Python 中的點陣圖和向量方法有效地檢測和勾畫一組分散的 2D 點中的孔？

此 Python 程式碼提供了類似於上一個答案中描述的點陣圖向量方法的實作。它透過計算位圖中的資料密度、識別未使用的區域、分割輸出並對結果進行多邊形化來尋找一組 2D 點中的孔。

import numpy as np
import cv2
import matplotlib.pyplot as plt

def find_holes(points, resolution=100):
    """Find holes in a set of 2D points.

    Args:
        points: A list of (x, y) tuples representing the points.
        resolution: The resolution of the bitmap to use. Higher resolution
            results in more accurate results, but is slower.

    Returns:
        A list of (x, y) tuples representing the vertices of the holes.
    """
    # Create a bitmap of the points.
    bitmap = np.zeros((resolution, resolution), dtype=np.uint8)
    for point in points:
        x, y = point
        bitmap[int(y * resolution), int(x * resolution)] = 255

    # Compute data density in the bitmap.
    density = cv2.dilate(bitmap, np.ones((3, 3))) - cv2.erode(bitmap, np.ones((3, 3)))

    # Identify unused areas in the bitmap.
    unused_areas = np.where(density == 0)

    # Segment the unused areas.
    segmented_areas = cv2.watershed(density, np.zeros((resolution, resolution), dtype=np.int32), markers=unused_areas[0], mask=bitmap)

    # Polygonize the segmented areas.
    holes = []
    for i in range(1, np.max(segmented_areas) + 1):
        mask = np.zeros((resolution, resolution), dtype=np.uint8)
        mask[segmented_areas == i] = 255
        _, contours, _ = cv2.findContours(mask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
        for contour in contours:
            hole = []
            for point in contour:
                x, y = point[0]
                hole.append((x / resolution, y / resolution))
            holes.append(hole)

    return holes


# Example usage

# Generate a set of random points.
points = [(np.random.rand(), np.random.rand()) for _ in range(1000)]

# Find the holes in the set of points.
holes = find_holes(points, resolution=50)

# Plot the points and the holes.
plt.scatter([x for (x, y) in points], [y for (x, y) in points], s=1, c='black')
for hole in holes:
    plt.plot([x for (x, y) in hole], [y for (x, y) in hole], c='red')
plt.show()

此程式碼使用 OpenCV 進行點陣圖操作，使用 NumPy 進行陣列操作操作，以及用於繪圖的 matplotlib。它可以輕鬆修改以使用不同的資料類型和座標系。

以上是如何使用 Python 中的點陣圖和向量方法有效地檢測和勾畫一組分散的 2D 點中的孔？的詳細內容。更多資訊請關注PHP中文網其他相關文章！