Tensorflow-Image-Classifier-Web-Demo

"Calorie Counter", helps you stay fit , by using computer vision to find which item are you eating and giving you nutrition facts

###End result (It's a video , click it)

Requirements

Web Server of your choice (nginx/apache/etc)(PHP must be installed)(See Any tutorial for setting up a webserver)

Files, What do they do?

uploads = user submitted images go here

index.html = upload site

postdata.php = main page, processing

styles.css = CSS styles for the web page

###Inspiration: Source: SAO Ordinal Scale - A1 Pictures

###First Step Clone this project (In terminal)

git clone https://github.com/AxelAli/Tensorflow-Image-Classifier-Web-Demo.git
cd Tensorflow-Image-Classifier-Web-Demo

then

###Setting a tensorflow image recognizer Follow this project.

https://github.com/AxelAli/Tensorflow-Image-Classification

When finished

cd ..

(We should now be in this proyect folder)

###Usage Open your webserver ####Click Choose file (in mobile you just open the camera)

####Select your picture

####Click Upload

####Results are in!

#####Orange is the image that you uploaded #####Purple is the name of the food and serving size #####Green is the amounts of each item

Known Bugs / Errors / Improvements

This should be a single webpage

if it doesnt find anything it shows a random item (we should check Accuracy of <80%)

data is hardcoded, this should be made with a database

if it doenst know which item it is, it should ask the user what is it and run the code to add it automatically