Before we can begin to count the number of fingers that the user is holding up in front of the camera, we need to verify that there is indeed a hand present in the frame. One of the methods we plan to investigate to do this is called a Cascade Classifier.

In a cascade classifier, there are multiple feature classifiers in series that work together to determine whether a certain object is within the input image. If the image passes through all of the classifier stages, then the image is said to contain the target object. The accuracy of the cascade classifier can be increased by increasing the number of stages as well as having a larger number of “positive” and “negative” training images with which to train the classifier. “Positive” images are those which contain the feature, and “negative” ones do not.

​

​

​

​

​

​

​

​

Another method which we could employ to detect the hands in the image is the ACF Object Detector. 

 

The ACF Object Detector functions by using Aggregate channel features (ACF) is a variation of channel features, which extracts features directly as pixel values in extended channels without computing rectangular sums at various locations and scales.Common channels include the colour channels, such as grey-scale and RBG, but many other channels can be encoded, depending on the difficulty of your problem (e.g. gradient magnitude and gradient histograms). ACF has advantages, such as a richer representation, accelerated detection speed and more accurate localisation of objects in the images when used in conjunction with a boosting method.

​

Training Example:

​

 

 

​

​

​

​

​