Image recognition is one of the more successful areas of machine learning. While most machine learning tasks require a relatively long period of time before their true benefits can be measured and quantified, image recognition is a familiar subject area that can be readily understood. In essence, it involves identifying objects and correctly categorizing them. It has several applications, ranging from identifying license plate numbers to face recognition, and is available in mobile devices and robotics.

OpenCV provides a standard interface for various image recognition tasks, and can also leverage hardware acceleration features to optimize performance.

Other well-known machine learning software for image processing include Caffe, cuDNN, TensorFlow, and others. Note that these packages are not limited to simply image recognition, but can be also used for other deep learning use cases.