OpenCV Combining SURF with Neural Network

Question

I want to recognize Vehicles(Cars, Bikes etc.) from a static image. I was thinking of using SURF to get me useful keypoints and descrip

Answer 1

An easy way to separte the object detected is runing an contour detector in the input image like that.

After its you can use the x,y coords from each key points associate to each feature vector recognized by the neural network,and check how many of these key points are in side each contour of each objects,

At the same time its let you put an treshold to validate a correct car detection, for example if you have 2 taxis;2 contour belong to each car and affter check how many key ponits are in each contour;

• contour belong taxi 1 have 20 key ponits inside
• contour belong taxi 1 have 5 key ponits inside

you can give as taxi 1 o car 1 as object recognized

On time you have the a validate object recognized and its contour,you can calcuialte the bouding box that closed the object

Another way to do it is to each training image extract its contours belong to each object ,calculate its bounding box use this like pure image to extract the features and make it to al picures en trainin set