Car Recognition

Question

Hi guys. I have started doing a project and i have to recognise how many cars are there in one photo. I started writing some code about it but i have a problem because i can recognise one car. Have you any ideas what could i do to make my code recognise more objects of the same type (cars) in one photo. I dont need codes i am just asking for ideas.

Answer 1

the best recipe for that in python is
OpenCV , NumPy,If You Want To Play Around With Pygame.Camera :p

Answer 2

Artificial intelligence, image processing, pattern recognition, neural network, great topic ^^

So your program learned to recognise one pattern. First question: Is the other object looking different and because of that not recognised?

Or is it a more general problem like the program stops scanning but shouldn't?

Answer 3

I have a ready solution for that, a convolutional neural network. The problem is - can you train it with a sufficient number of "good" examples? Do you have a clean dataset you can use?

Answer 4

ty guys my haar cascade needed some training . now i need it to detect cars at night any ideas?

Answer 5

also do you think neural networks is better or haar cascade?

Answer 6

you could create a list and every time you find a car take a look to the list
if the list contains the car you keep looking for more cars until you have searched the full photo

something like this (i dont know python)
list cars
while(searchedOnWholePhoto) {
if(foundACar) {
if(!cars.contains(car){
cars.add(car)
}
}
}