深度學習python之用Faster-rcnn 檢測結果（detections.pkl 和class_pr.pkl） 在原圖畫出box

class_pr.pkl 包括召回率（Recall），精确率（Precision），平均正确率（Average_precision(AP) ）

在Spyer中敲入如下代碼即可打開，在變量欄檢視具體數值（本文隻對一類目标進行檢測）

import cPickle as pickle  
f = open('ship_pr.pkl')  
info = pickle.load(f)  
           

下面是使用detections.pkl畫框，裡面包括的是list，list裡包含兩個list，坐标值在第二個list裡

import os
import os.path
import numpy as np
import xml.etree.ElementTree as xmlET
from PIL import Image, ImageDraw
import cPickle as pickle 

f = open('ship_pr.pkl')  
ship_pr = pickle.load(f) 

test_file = 'test.txt'
file_path_img = 'JPEGImages'
save_file_path = 'pkl/results'

with open(test_file) as f:
    image_index = [x.strip() for x in f.readlines()]

f = open('detections.pkl')  
info = pickle.load(f)


dets = info[]

num = 
for idx in xrange(len(dets)):
    if len(dets[idx]) == :
        continue
    img = Image.open(os.path.join(file_path_img, image_index[idx] + '.jpg')) 
    draw = ImageDraw.Draw(img)
    for i in xrange(len(dets[idx])):
        box = dets[idx][i]
        draw.rectangle([int(np.round(float(box[]))), int(np.round(float(box[]))), 
                    int(np.round(float(box[]))), int(np.round(float(box[])))], outline=(, , ))
    img.save(os.path.join(save_file_path, image_index[idx] + '.jpg'))