Research On Dual View Visual Measurement Technology Of Bottom Loading Canister

Large-scale laser devices need to be replaced with a large number of optomechanical modules,also known as Line Replaceable Units(LRU modules).The LRU module is mainly responsible for a series of laser experiments.The response mechanism requires high precision and cleanliness.Special equipment is required for transshipment and disassembly.About two-thirds of the modules need to go from bottom to top.The school is installed,which is the download module.The gap between the top of the transfer car module and the bottom of the mounting position is very small,and the conventional visual measurement method is difficult to implement.In this paper,the positioning and docking of the following optical module module is the research object,and a dual-view visual measurement method is proposed,which is mainly applied to the narrow viewing angle and short viewing distance environment.Firstly,this paper analyzes the project requirements of the dual-view visual measurement system,and draws the workflow,that is,when the loading system is installed,when the transport vehicle travels to the designated area,the dual-view visual measurement system is used to identify the large laser.The identification feature block installed at the bottom of the clean room is extracted,and the coordinates of the feature point are extracted and converted by coordinates to obtain the rotational offset of the transfer box and the clean room.Then introduces the camera calibration principle and the traditional binocular camera calibration principle,which leads to the calibration method of dual-view visual measurement.At the same time,the mathematical formula of coordinate transformation after acquiring the central coordinate of the identification feature block is derived.Secondly,this paper proposes a logarithmic transformation image enhancement based on bilateral filtering for narrow viewing angle and short-sighted environment.It is compared with histogram equalization enhancement and top hat morphology enhancement.analysis.This method is more obvious than other methods,and it is also robust to normal lighting conditions.Then,in order to solve the problem of excessive time consumption in hough transform circle detection,a round detection method based on random hough transform is proposed,which significantly improves the efficiency of circle detection and preserves the high precision of hough transform circle detection.Finally,the design and construction of the dual-view visual measurement experimental platform is carried out.A special extended calibration plate is designed to verify the calibration of the dual-angle camera.The calibration result re-projection error is less than 0.1,which is similar to the traditional binocular camera calibration result.At the same time,the effectiveness of the overall vision measurement system was verified.The final result shows that the accuracy of the dual-view visual measurement meet the actual needs,and the requirements of the fully automatic closed-loop real-time docking control are verified,which has a guiding effect on the subsequent engineering implementation.