Machine Learning: Image Classification

Published on Sep 19th, 2019 10:21 PM

task one is image classification. it sounded easy until i started.

This particular taskl was about classification of two sets of data and training the program to know and recognize which data is a car plate number and which is not.

During this process, i gathered images of various car plates both from my partner Google, and the ones i took with my Smartphone. i got some random pictures from my desktop.


I grouped the two data set into folders with different name.

i begand building the classifier with Google Colab, Then imported several python libraries eg: pandas, numpy etc. then i mounted the two data folder into my Google drive for easy assessment.

Logistic Regression:

(this is from statistics)

i used Logistics Regression model which helped to resize all the images. Then I used Logistics regression class from the sklearn module and put the data into the model.

I also created a function that predicts if an image is a plate number or not..

a few challenges i had was lack of regular power supply, Insufficient mobile data.

I encountered a lot of errors; namely os error, fileNameError,