Instagram, a popular photo-sharing platform, has millions of users, with 70 Million photos uploaded on average per day. Users provide information about images by putting "hashtags" on them. The metadata also contains geographical location where the picture was captured. However, user-provided hashtags can sometimes be imprecise or irrelevant. It is important to visually recognise the image content so it may be correctly analysed with respective metadata in order to get interesting collective or user-specific statistics. This book deals with the problem of visual recognition of Instagram photos (particularly food images) and aims to solve it by classification via supervised learning mechanisms. The book provides guide to creating standalone and distributed versions of the proposed solution along with code snippets to aid the programmers. The tools include Apache Spark Machine Learning library and OpenCV. The proposed solution can recognise food content with up to 84% precision and 94% accuracy, and can be integrated with image streams to get live statistics regarding popular foods in a specific region.