解释 NMS (Non Maximum Suppression) 算法的原理

def non_max_suppression(boxes, scores, iou_threshold):
    # boxes: (N, 4) array of [x1, y1, x2, y2] coordinates
    # scores: (N,) array of confidence scores
    # iou_threshold: IoU threshold for suppression

    # 获取按置信度分数降序排列的索引
    order = scores.argsort()[::-1]
    keep = []

    while order.size > 0:
        # 选择置信度分数最高的边界框
        i = order[0]
        keep.append(i)

        # 计算当前最佳检测框与其他候选框的IoU
        xx1 = np.maximum(boxes[i, 0], boxes[order[1:], 0])
        yy1 = np.maximum(boxes[i, 1], boxes[order[1:], 1])
        xx2 = np.minimum(boxes[i, 2], boxes[order[1:], 2])
        yy2 = np.minimum(boxes[i, 3], boxes[order[1:], 3])

        w = np.maximum(0.0, xx2 - xx1)
        h = np.maximum(0.0, yy2 - yy1)
        inter = w * h

        area = (boxes[order[1:], 2] - boxes[order[1:], 0]) * (boxes[order[1:], 3] - boxes[order[1:], 1])
        iou = inter / (area + (boxes[i, 2] - boxes[i, 0]) * (boxes[i, 3] - boxes[i, 1]) - inter)

        # 抑制IoU超过阈值的边界框
        inds = np.where(iou <= iou_threshold)[0]
        order = order[inds + 1]

    return keep