The Research Of Pump Truck Anti Collision Algorithm Based On Hybrid Bounding Box

With the rapid development of social economy and the emergence of a variety of large-scale construction projects, the concrete pump truck is widely used in many fields.Especially in some difficult project, the concrete pump truck is becoming an important equipment with its flexible characteristics in modern engineering projects. But in the actual operation, due to the interference external factors and different operating personnel experience, in operation process, the accident will be most likely appear due to a variety of false positives and false operation.Firstly, this paper put the dual camera installation in the boom tip top of pump truck,samples the image around the obstacles based on image acquisition method of binocular camera, which can calculate the obstacle distance, profile and other parameters by binocular stereo vision measurement method, using MATLAB to analysis the performance of distance measurement and constructs the three-dimensional model of concrete pump truck around obstacles for the pump vehicle anti collision algorithmSecondly, for the problem of low efficiency, low real-time performance and the large redundancy of the bounding box in the sphere bounding box. The paper has proposed the anti collision detection algorithm based on Genetic Optimization and multi layer bounding box, which improves the operation speed of the algorithm. In order to solve the bounding box compactness problem, this paper proposes a anti collision detection algorithm based on Genetic Optimization and hybrid bounding box,which use different types of bounding box surrounded by multi-level of pump truck boom and different kinds of obstacles and the improved algorithm. The simulation results show that, the improved algorithm has better real-time and bounding box compactness.