in recent years the popularity of quadrotor unmanned aerial vehicles (uavs) has increased. today, uavs are widely used by military and police forces for surveillance. they are used by industry for such tasks as traffic monitoring, infrastructure inspection or even delivery of goods. they are used by individuals for hobby flying and aerial photography. it is currently of great interest in the research community to improve the level of autonomy of the uav for these and future uses. one particular problem is the ability to stabilize over and land on a moving platform. this situation can easily arise for a quadrotor returning to a ship at sea or even a landing pad affixed to a vehicle. many current techniques rely on knowledge of the platform and its motion, or a predictive model. this information is not always available or accurate. a solution that does not require knowledge of the target is desirable. this thesis deals with practical implementation of optical flow based position stabilization and autonomous landing algorithms for a quadrotor uav. the quadrotor used is a common low cost platform with a large open source community. firstly, non-linear estimation and control techniques are implemented for the attitude stabilization using low-cost sensors and limited computational power. some methods for the system parameters estimation are presented and some challenges related to the implementation are discussed. despite the ability of the attitude controller to stabilize the orientation of the quadrotor, hovering and landing precisely over a specific area is not possible without a position stabilization scheme. in applications where gps signals are not available and the hovering target is a priori unknown, it is common to rely on visual information. in this context, this thesis aims for the development of an efficient optical-flow-based position stabilization and autonomous landing scheme for the quadrotor uav.