map（mean Average Precision)计算及代码实现

参考文献:https://github.com/rafaelpadilla/Object-Detection-Metrics  #强烈推荐这个

https://www.imooc.com/article/44040

https://blog.csdn.net/bismarckczy/article/details/86520342  #这个的计算方式不懂

https://www.jianshu.com/p/82be426f776e

目标检测中最常用的测评指标就是map(mean average precision),介绍其计算方法之前，先介绍几个相关概念：

IOU(Inetersection over union):交并比

TP（True Positive）：=number of detections with iou>0.5，指正确预测的bbox

FP(False Positive)=number of detections with iou<=0.5 or detected more than once,指预测错误的bbox