yolov8iou改进

YoloV8的改进策略主要是将CIoU替换成Wise-IoU，支持EIoU、GIoU、DIoU、SIoU无缝替换。其中，CIoU是传统的IoU计算方法，而Wise-IoU是一种新的IoU计算方法，它可以更好地处理物体之间的重叠情况，从而提高检测精度。此外，YoloV8还对模型的骨干网络进行了改进，采用了更加高效的CSPDarknet53网络，从而提高了模型的检测速度和精度。

下面是YoloV8的改进策略的代码实现：

# 使用Wise-IoU计算IoU
def wise_iou(box1, box2):
    # 计算交集的左上角和右下角坐标
    x1 = max(box1[0], box2[0])
    y1 = max(box1[1], box2[1])
    x2 = min(box1[2], box2[2])
    y2 = min(box1[3], box2[3])
    # 计算交集面积
    inter_area = max(0, x2 - x1 + 1) * max(0, y2 - y1 + 1)
    # 计算并集面积
    box1_area = (box1[2] - box1[0] + 1) * (box1[3] - box1[1] + 1)
    box2_area = (box2[2] - box2[0] + 1) * (box2[3] - box2[1] + 1)
    union_area = box1_area + box2_area - inter_area
    # 计算IoU
    iou = inter_area / union_area
    # 计算中心点的距离
    center_x1 = (box1[0] + box1[2]) / 2
    center_y1 = (box1[1] + box1[3]) / 2
    center_x2 = (box2[0] + box2[2]) / 2
    center_y2 = (box2[1] + box2[3]) / 2
    center_distance = (center_x1 - center_x2) ** 2 + (center_y1 - center_y2) ** 2
    # 计算Wise-IoU
    wise_iou = iou - center_distance / union_area / 4
    return wise_iou