python统计分析darknet分析指令结果

cmd下运行darknet指令

darknet classifier test metal.data 19448.txt backup/19448_last.weights > ab1009.txt

在文件夹下打开ab1009.txt

seen 64 
E:/MyCode/fushikang2019/ImgOCT/normal/171222141438_00002/s064_normal.jpg	0.365515	0.634485
E:/MyCode/fushikang2019/ImgOCT/abnor/171222103915_00011/s006_abnor.jpg	1	3.52733e-09
E:/MyCode/fushikang2019/ImgOCT/normal/171224091139_00012/s122_normal.jpg	0.999735	0.000264592
E:/MyCode/fushikang2019/ImgOCT/normal/171222084404_00012/s072_normal.jpg	0.00825327	0.991747
E:/MyCode/fushikang2019/ImgOCT/normal/171222103742_00012/s091_normal.jpg	0.999897	0.000103443
E:/MyCode/fushikang2019/ImgOCT/normal/171224091022_00002/s056_normal.jpg	0.210825	0.789175
E:/MyCode/fushikang2019/ImgOCT/normal/180102154232_00012/s122_normal.jpg	0.99576	0.00424046
E:/MyCode/fushikang2019/ImgOCT/abnor/171223101358_00011/s160_abnor.jpg	0.999989	1.1315e-05
E:/MyCode/fushikang2019/ImgOCT/normal/171224153637_00011/s015_normal.jpg	1	9.10471e-08
E:/MyCode/fushikang2019/ImgOCT/abnor/171226150449_00002/s005_abnor.jpg	0.946796	0.0532044
E:/MyCode/fushikang2019/ImgOCT/abnor/171229090713_00011/s157_abnor.jpg	0.999996	4.39967e-06
E:/MyCode/fushikang2019/ImgOCT/normal/171229100929_00012/s017_normal.jpg	0.0263349	0.973665
E:/MyCode/fushikang2019/ImgOCT/normal/171224084004_00011/s032_normal.jpg	1	1.02775e-07
E:/MyCode/fushikang2019/ImgOCT/normal/171231094310_00012/s126_normal.jpg	0.76249	0.23751
E:/MyCode/fushikang2019/ImgOCT/normal/171229094942_00002/s006_normal.jpg	0.999643	0.000357216
E:/MyCode/fushikang2019/ImgOCT/normal/171225151905_00002/s001_normal.jpg	0.0452934	0.954707
E:/MyCode/fushikang2019/ImgOCT/abnor/171225180635_00011/s061_abnor.jpg	0.999684	0.000315564
E:/MyCode/fushikang2019/ImgOCT/normal/171221150724_00012/s042_normal.jpg	0.0476515	0.952349
E:/MyCode/fushikang2019/ImgOCT/normal/171225151411_00011/s122_normal.jpg	0.999923	7.67407e-05
E:/MyCode/fushikang2019/ImgOCT/normal/171222150056_00004/s026_normal.jpg	4.53133e-08	1
E:/MyCode/fushikang2019/ImgOCT/abnor/171222163513_00011/s082_abnor.jpg	0.999983	1.67547e-05

任务说明:

分别t统计abnor和normal两种类别图片中,识别错误的图片和识别正确的图片

import re

mf = open('ab1009.txt', 'r')
normal_error = []
normal_correct = []
abnor_correct = []
abnor_error = []
num = 0
for line in mf:
    #print(line)
    m = re.findall(r"\t\d+\.?\d*e?-?\d*", line)
    if 'normal' in line:
        #print(m)
        if len(m) > 1:
            #print(m[0])
            #print(m[1])
            if float(m[0]) < float(m[1]):
                #print('1')
               # print(line)                
                normal_error.append(line)
            if float(m[0]) > float(m[1]):
                normal_correct.append(line)
    if 'abnor' in line:
        if len(m) > 1:
            if float(m[0]) < float(m[1]):
                #print(line)
                abnor_correct.append(line)
            if float(m[0]) > float(m[1]):
                print(line)
                num = num+1
                print(num)
                abnor_error.append(line)
mf.close()
mfile1 = open('normal_error.txt', 'w')
for aa in normal_error:
    mfile1.write(aa)
mfile1.close()
mfile2 = open('normal_correct.txt', 'w')
for aa in normal_correct:
    mfile2.write(aa)
mfile2.close()
mfile1 = open('abnor_error.txt', 'w')
for aa in abnor_error:
    mfile1.write(aa)
mfile1.close()
mfile1 = open('abnor_correct.txt', 'w')
for aa in abnor_correct:
    mfile1.write(aa)
mfile1.close()

该代码会保存四个文件:

normal_correct.txt,normal_error.txt,abnor_correct.txt,abnor_error.txt

其中normal和abnor表示该图片文件现在标注为normal或abnor

correct和error表示识别正确或错误,例如标注为normal识别依然为normal则说明识别正确。

normal_error.txt文件示例如下:

E:/MyCode/fushikang2019/ImgOCT/normal/171222141438_00002/s064_normal.jpg	0.365515	0.634485
E:/MyCode/fushikang2019/ImgOCT/normal/171222084404_00012/s072_normal.jpg	0.00825327	0.991747
E:/MyCode/fushikang2019/ImgOCT/normal/171224091022_00002/s056_normal.jpg	0.210825	0.789175
E:/MyCode/fushikang2019/ImgOCT/normal/171229100929_00012/s017_normal.jpg	0.0263349	0.973665
E:/MyCode/fushikang2019/ImgOCT/normal/171225151905_00002/s001_normal.jpg	0.0452934	0.954707
E:/MyCode/fushikang2019/ImgOCT/normal/171221150724_00012/s042_normal.jpg	0.0476515	0.952349
E:/MyCode/fushikang2019/ImgOCT/normal/171222150056_00004/s026_normal.jpg	4.53133e-08	1
E:/MyCode/fushikang2019/ImgOCT/normal/171222174148_00004/s014_normal.jpg	0.109819	0.890181
E:/MyCode/fushikang2019/ImgOCT/normal/171225151317_00012/s074_normal.jpg	0.00852944	0.991471

 

  • 0
    点赞
  • 0
    收藏
    觉得还不错? 一键收藏
  • 打赏
    打赏
  • 0
    评论
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包

打赏作者

haimianjie2012

你的鼓励将是我创作的最大动力

¥1 ¥2 ¥4 ¥6 ¥10 ¥20
扫码支付:¥1
获取中
扫码支付

您的余额不足,请更换扫码支付或充值

打赏作者

实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值