Image Processing with Backpropagation algorithm

First of all, I don't have the code yet for this project. I just want to ask first what would be the good approach/way to do this project before I start with the code (although I'm not that good in MATLAB). So our project is to determine the shades of green of the leaves and the shape of the diseases present on the leaves. What we have proposed so far is to make a database for the shape and color using the backpropagation. And then the test image will be compare to the database. Do you think that's a good approach? Can you suggest some other approach to do this better.

Thank you and God bless :)

Answer 1

[Greg Heath]

Backpropagation is not used to directly create a data base.

However, if you have a data base of inputs and targets, one of the backpropagation functions like fitnet (regression or curvefitting) or patternnet (classification or pattern recognition) is used to NOT ONLY output close approximations to training target vectors when the corresponding training input vectors are presented, BUT, more importantly, generalize to nontraining data.

Backpropagation can be used to create nets for testing whether or not your choice of targets and target coding (real?, max? min? binary? unity sum? etc) is useful.

Hope this helps.

Thank you for formally accepting my answer

Greg

[Elvin]

Thank you for your answer. That helps a lot. I thought I could use backpropagation for the database. May I ask how am I going to create the database for the shape and color? THanks

[Greg Heath]

You should post a new question on how to implement image feature extraction in order to create a data base of shape and color features.

Answer 2

[Shashank Prasanna]

Hi Elvin, what you are proposing is a supervised learning approach. backpropogation (Neural Networks) to train your data is one good approach. You could also explore other supervised learning approaches in the statistics toolbox:

http://www.mathworks.com/help/stats/supervised-learning-machine-learning-workflow-and-algorithms.html

Because of the convenient way the functions have been written, its easy to just try different algorithms quickly that saves you time.

Here are some good ways to get started:

http://www.mathworks.com/discovery/machine-learning.html

[Elvin]

Thank you for your reply sir. I appreciate it. :)

By the way, I forgot to mention that we already proposed our approach and our panels already approved it. I'm just confuse with our process. So is it right that I should make the color and shape database using the backpropagation algorithm?

[Shashank Prasanna]

backpropogation can be used to train on a dataset for future prediction and is a popular approach. You can use the Neural Network Toolbox to do that.

It is not clear what you mean when you say "make the color and shape database"

[primrose khaleed]

i want create database of image and i want to enter this image into NN..my asking which image saving in database ?? the original images or after fuature extraction?

Answer 3

[Greg Heath]

You need to search on image feature extraction.

[Image Analyst]

And when you do, attach an image.

Answer 4

[Elvin]

Any help with this one? Thanks