| In recent years,the photographing robot,which combines the robot technology with the specific needs of the photography industry,has greatly saved the human cost,improved the shooting efficiency,and optimized the shooting effect of the scene or target to a considerable extent.It is gradually extended to the film production,the live broadcast of the competition and the photography.For the indoor photographing and camera shooting,there is still a lack of necessary researches about the closer interaction between the robot and photographers' needs,or the robot and the actual state of the shooting target,or the robot and the photographing environment.There are also some inevitable problems in the existing products,such as low adaptability and low-level intelligence.This paper is based on the theory of the visual servo control of the Omnidirectional mobile robot and the detection and tracking of the target,esspecially combined with the basic theory about photography and camera shooting.In order to achieve the given control requirements and expected shooting effect,the mobile robot shooting control system related to the actual state of the target in two common shooting modes is designed respectively.First,the omnidirectional moving platform of the robot is designed,and then the forward and inverse kinematics model is analyzed.Based on the kinematics model,the control system is designed for controlling platform's position and trajectory tracking.The simulation results verify the feasibility of the kinetic control system.And the programs for controlling the motors' speed are also designed.Secondly,the robotic control scheme of fixed-point shooting is designed.The platform is controlled by the position-based visual servo control mode.The method of selecting and extracting feature points is given.The improved method for the P3 P problem in the pose estimation is presented,and whose rationality is verified by the simulation result.The method of solving the motion parameters during the servo control is presented,and the patterns of basic picture composition about photography and camera shooting are analyzed.Based on the composition of the object centered,the static shooting and adjusting algorithm of the pan-tilt-zoom(PTZ)camera is designed.Thirdly,the control scheme of the shooting during tracking the target is designed.The dynamic tracking model of the platform is deduced and constructed.The image-based visual servo control system of the mobile platform is designed.The feature points are extracted and obtained in the corresponding image space based on the Kinect camera,and the approach for depth estimation is studied.The control effect under the combination of different gain values is discussed.The effectiveness of the control system is verified by simulation.For the tracking photography,an improved shootingalgorithm based on the existing static shooting algorithm is proposed,which is fused with Kalman filtering and Camshift algorithm.And and the flow chart of the algorithm is given.Finally,the experiment platform of the photographing robot is built,and the motion control interface of the platform is designed.The specific factors affecting the shooting effect are discussed,and the experiments are carried out on the two kinds of shooting methods respectively.The experimental results show that the control effect of the platform and the shooting effect of the PTZ camera meet the expected index of the experiment,and the schemes of the control system and the algorithms which are proposed in this paper are reasonable. |
PDF Full Text Request