【图像】手写数字识别 k近邻

 

1 数据

                            

--train_npy  训练数据     --test_npy 测试数据

.npy格式，01二值，尺寸32*32

数字包含0~9 ，文件名第一位为标签。数据不多，可以自己再搞一些

链接：https://pan.baidu.com/s/1tB-B5v1eQbYUwh1808bZFA 
提取码：lucy 

2 代码

from numpy import *
import numpy as np
import cv2
from os import listdir
import operator

def classify0(inX, dataSet, labels, k):
    dataSetSize = dataSet.shape[0]
    diffMat = tile(inX, (dataSetSize,1)) - dataSet
    sqDiffMat = diffMat**2
    sqDistances = sqDiffMat.sum(axis=1)
    distances = sqDistances**0.5
    sortedDistIndicies = distances.argsort()
    classCount={}
    for i in range(k):
        voteIlabel = labels[sortedDistIndicies[i]]
        classCount[voteIlabel] = classCount.get(voteIlabel,0) + 1
    sortedClassCount = sorted(classCount.items(), key=operator.itemgetter(1), reverse=True)
    return sortedClassCount[0][0]



def handwritingClassTest():
    Labels=[]
    trainList=listdir('train_npy')
    testList =listdir('test_npy')
    train_num=len(trainList)
    test_num = len(testList)

    trainMat=np.zeros((train_num,1024))
    for i in range(train_num):
        fileNameStr=trainList[i]
        trainMat[i, :]=np.load('./train_npy/{}'.format(fileNameStr))
        fileStr=fileNameStr.split('.')[0]
        num=int(fileStr.split('_')[0])
        Labels.append(num)
    right_count=0

    for i in range(test_num):
        fileNameStr=testList[i]
        cur_testMat= np.load('./test_npy/{}'.format(fileNameStr))
        fileStr=fileNameStr.split('.')[0]
        num=int(fileStr.split('_')[0])
        classifierresult=classify0(cur_testMat,trainMat,Labels,3)
        print("get:%d,real:%d"%(classifierresult,num))
        if (classifierresult==num):
            right_count+=1
    print("正确率:{}%".format(np.round(right_count*100/test_num,2)))

if __name__ == '__main__':
    handwritingClassTest()

3 输出