Image-Classifier-by-alexnet

It is a Image classifier by using alexnet architecture for classifying one object in a image with high accuracy.

Instructions

1. If you have images then paste it in photos\1 folder for 1 st image type and photos\2 for 2 nd and so on you can create more folders named '4' , '5', and so on depending upon number of image types you have or how much types of images you want to classify

2. You can run imagesfromvideo.m to get images from live web cam

3. Then run Imageclassifier.m and mention the number of image types you want to classify there of it will take about 3-10 min of training depending on how much data you have and paste the path of you new image and test it .

Note: Don't change the folders name 1,2,3,.. in photos folder write the names of image classes in the code and It can only detect only one object in an image.