Classification of objects from a video ( human, animals, others(cars etc.,) )

Question

Hi, I am new to image processing field. My project is to classify the objects in an image/Video. The input is an image/video from a surveillance camera. I should classify the objects into three classes. 1) Humans 2) Animals 3) Others (cars). We can consider a fixed background.

Can anyone suggest any proven algorithms or papers or opensource code for this purpose? of course, googling is an obvious solution but always an experienced eye is always helpful!

Answer 1

The classification of various objects is the ability to learn the type of object.  So rather than merely recognizing that something is there, classification determines the type of object: human, animal, car, truck, boat. 

Two approaches are possible:

1 - Use of Bag of Words. Usually, they have four steps:

• Get the feature points from training data;

• Create a vocabulary from the features;

• Train (using SVM for example);

• Test in an image (not in training examples).

There is an example here and in opencv/samples/cpp/bagofwords.cpp

2 - Use of a latent SVM detector. The explanation is here. Have a running example of detection/categorization in opencv/samples/cpp/latentsvmdetector.cpp. You also require the opencv_extra/testdata/cv/latentsvm to properly test it. 

For more information regarding Human Motion Detection and Tracking for Video Surveillance, refer to the following link:

https://pdfs.semanticscholar.org/11cd/be81c995377edfb943d652d3a5c22f859f06.pdf