i need an idea to start my project

Question

1.00/5 (2 votes)

See more:

I'm creating a system that analyze the car accidents pictures using C# windows application

im not good at artificial intelligence

The system based on sending a picture of the accident to the program by MMS , and then the program analyzes it and decide who's the guilty ( sorry i don't know the proper word ), so we can estimate the claim by using the system , So any idea how can i implement

this idea ?

Posted 16-Feb-12 7:39am

Hosain.ahm

Add a Solution

Comments

ZurdoDev 16-Feb-12 13:52pm

You'll need to find some 3rd party software that can analyze images.

Dave Kreskowiak 16-Feb-12 13:56pm

Sooo, what "pictures of the accident" are you going to show this system?? What are these pictures of??

Truthfully, based on your very brief description, I don't see how you're going to pull this one off unless you're a GOD in Machine Vision and Object Recognition Systems.

Johannes Hillert 16-Feb-12 20:28pm

You definitely won’t be able to determine who was responsible for the car accident from a picture.. no chance. You would need much more information and some kind of simulation how the accident happened. I have to ask it.. did you think about a possible implementation or the question at all?

2 solutions

Add a Solution

Add your solution here

Treat my content as plain text, not as HTML

Preview 0

…

Existing Members

Sign in to your account

...or Join us

Download, Vote, Comment, Publish.

Your Email
Password
Forgot your password?

Your Email
This email is in use. Do you need your password?
Optional Password

I have read and agree to the Terms of Service and Privacy Policy
Please subscribe me to the CodeProject newsletters

When answering a question please:

Read the question carefully.
Understand that English isn't everyone's first language so be lenient of bad spelling and grammar.
If a question is poorly phrased then either ask for clarification, ignore it, or edit the question and fix the problem. Insults are not welcome.
Don't tell someone to read the manual. Chances are they have and don't get it. Provide an answer or move on to the next question.

Let's work to help developers, not make them feel stupid.

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

krumia · Answer 1 · 2012-02-16T16:34:00

This is HUGE. What can I say: You have chosen this as a project only because of your lack of knowledge in AI/Machine Vision.

If you're trying to do a project for a degree, cut down the scope. Think of something simple. Like "Differentiating vehicles in a highway based on static images". Even that should take something like one year (full time) to complete.

And really, if your not good at AI, abandon this. And choose something that can be done using what you're good at. You don't get an A+ because you chose an AI project. You will get an A+ if you complete whatever the project you're doing, and present it nicely in the final presentation, and defend it in a impressive way.

Good luck.

Dr.Walt Fair, PE · Answer 2 · 2012-02-16T08:03:00

Solution 1

I would start by talking to an insurance adjuster and trying to figure out what they look at when writing up a claim. Then, once you understand how the claim should be estimated, try to figure out how to do that based on your data (images). Only after you understand what needs to be done and how it relates to your input data (images), then you can start to design a system.

Good luck. I have a feeling you're going to need it.

Posted 16-Feb-12 8:03am

Dr.Walt Fair, PE

Comments

Hosain.ahm 16-Feb-12 14:17pm

i already did that , the point is how the program would understand that this car hit the other from the back , or from the front , ofcours i want this to be done based on the images. Do i need some feature extraction algorithms ?

Dave Kreskowiak 16-Feb-12 14:35pm

How are you going to recognize features that are unrecognizable when they've been crunched in a crash??

In reality, I don't think you've going to get this to work at all. Machine Vision technology is nowhere near the point where it can replace a cops and/or an adjusters eyes and brain.

You can show it a picture of the backend of a car that's smashed, but that's only a very small piece of a much larger vision puzzle.

How is the machine going to know the position of the car at the real world site?? How is it going to know WHEN the car was hit?? Was it in the initial impact or was it collateral damage from another victim car hitting it??

You're biggest problem right now isn't how to recognize a smashed car. It's how are you going represent the scene of the accident in data?? For example, there is severe damage to the front end of two cars, one on each side of a two-lane highway. Without knowning anything else, how are you going to determine who is at fault?? Who crossed the line into opposing traffic?? Maybe both of them did! Then what??

Seriously, I can easily see how this system would cost $100,000,000+ to develop.

Dr.Walt Fair, PE 16-Feb-12 14:28pm

Probably so, however, that is no trivial task!