Vision-Based Control of the Overhead Crane

Abstract: Through the years, cranes have been used extensively for many different tasks. For example construction work, shipbuilding and cargo transportation. One of the tedious tasks that cranes perform is loading and unloading containers. This is a process which can be optimized with the aid of control theory. While moving cargo, the crane operator must make sure that crane motion does not cause the cargo to accelerate too quickly, as the cargo may then start to oscillate. Such motion may disrupt unsecured cargo and put unnecessary stress on secured cargo. Today the crane movement is directly controlled by the crane operator; this however severely limits the crane speed due to human inability to control the crane optimally. This thesis presents and investigates a method of controlling the crane movement. The method utilizes computer vision to determine cargo position which is used to control crane movement optimally.