Computer Learning Practice (iii) KNN Handwriting Recognition System

Objective: To identify numbers 0-9 Using KNN

Material: 32*32 digital square arrays (saved as text files)

#-*-Coding: UTF-8-*-from numpy import * def img2vector (filename): # generate a 1*1024 array (zeros is a numpy function, as for the differences between array and list, we will not introduce them here.) returnvect = zeros () # use the open function to open a text file Fr = open (filename) # cyclically read the file content for I in range (32): # Read a row and return the string linestr = Fr. readline () for J in range (32): # Read string 0 or 1 returnvect [* I + J] = int (linestr [J]) # Return this array return returnvect

This program is very clear and I will not explain it any more. Let's take a look at how the classifier is implemented:

 

Computer Learning Practice (iii) KNN Handwriting Recognition System