System and method for human detection and counting using background modeling, HOG and Haar features

A method for detecting a human present in an image, the method comprising:

• acquiring, by utilizing instructions from memory that are executed by a processor, an image associated with the human, from a light detection sensor;
• capturing, by utilizing instructions from memory that are executed by the processor, a foreground of the image associated with the human using background modeling (BG);
• enhancing, by utilizing instructions from memory that are executed by the processor, a quality parameter of the image associated with the human around the foreground to generate a designated foreground;
• deriving, by utilizing instructions from memory that are executed by the processor, a histogram of an oriented gradient (HOG) classifier for the designated foreground of the image associated with the human, wherein the HOG classifier is derived to detect the human in the image associated with the human;
• deriving, by utilizing instructions from memory that are executed by the processor, a Haar classifier upon detecting the human, wherein the Haar classifier is derived to detect a face corresponding to each human present in the image associated with the human;
• and
• combining, by utilizing instructions from memory that are executed by the processor, the designated foreground generated using the BG, the HOG classifier, and the Haar classifier to perform a time domain analysis in order to detect presence of the human in the image associated with the human, wherein the time domain analysis indicates each pixel, associated the human, is analyzed over a period of time to determine a probability indicating the presence of the human detected in the image associated with the human.