User:Arduinolover/sandbox

Object class detection is a computer technology that deals with detecting objects of a certain class (such as humans, buildings, or cars) in digital images and videos. Well-researched domains of object class detection include face detection and pedestrian detection. Object class detection has applications in many areas of computer vision, including image retrieval and video surveillance.

It is used in face detection and face recognition. It is also used in tracking objects, for example tracking a ball during a football match, tracking movement of a cricket bat, tracking a person in a video.

As we know every object class has its own special features that helps us in classifying the class. For example all circles are round. It uses the special features of an object. For example if you are looking for circles your main concern will be to look for object that are at a particular distance from a point i.e. center. Similarly while looking for squares you will look for objects that are perpendicular at corners and are having equal side length. Similar things goes for faces. A face consists of eyes, nose, lips, and the features taken can be skin color, distance between eyes.

The advantage we are having is, an image is made of pixels. So in most cases we know the location of next point, it will be connected to our current pixel. Starting with circles, take an image, convert it to gray scale, and detect edges. Move along edges, draw normal, they will intersect at center. Do this for entire circle or find connected edges and calculate Euclidian distance between center and connected points. Another algorithm is move along connected edges rotation of tangent will be uniform, because of symmetry. So whenever there is an abrupt change in rotation, you are out of circle.
For squares, move along edges. First of all check if they are straight lines or not (check pixels are having either same x or y co-ordinates). After that look for a 90 degree change in angle(if you were moving along a horizontal line then at corner y co-ordinate will stop changing and x will start changing).

^{[1] [2]}

Multiple object class detection