Pattern Recognition is widely spread as a basis for classification purposes. However, the common pattern recognition and classification methods are usually used on non-spatially-correlated data. For example, moment invariants are a very basic and popular tool for this purpose. Such methods cannot be easily applied on flow data, like wind streams and water flows, magma flows, blood flows for medical purposes, air flow simulations for the design of cars and airplanes. For this reason, this book explains how classic moment invariants have been transformed to facilitate pattern recognition on flow data. The properties of the "Flow Moment Invariants" are analyzed and used to perform Feature-based and Comparative Visualization of flow data.