Research On The Cooperative Target Optimization For Monocular Vision Measurement

The monocular vision measurement system has a wide range of applications,and plays an important role in the aerospace field and th e civilian field.Its measurement accuracy has attracted much attention.In the monocular vision measurement system,there are many factors that affect the accuracy,such as solution model,camera focal length,camera calibration error,target parameters,and so on.At present,there are few studies on different target feature parameters.Through this topic,we study different target feature parameters and compare the effect of different feature parameters on the accuracy of the algorithm.In this paper,we study the non-coplanar object composed of multiple feature points.Comparing with the commonly used methods,we use multi-point non-coplanar objects for positioning,which can be linearly solved,avoiding complicated solution process and multiple solutions.The problem is more simple.In this paper,the non-coplanar target consisting of seven feature points is used as the target feature,and the position arrangement of point features is used as the target parameter.The position and orientation measurement algorithm model based on multi-point non-coplanar target is established and the pose and position measurement error model is established.Based on this,an objective function that optimizes non-coplanar target parameters is proposed.Through software programming,simulations are performed to solve the pose,and non-coplanar target parameters are optimized.The non-coplanar target in the experiment is composed of seven non-coplanar features,arranged according to the different positions of the points,and a total of five parameters are used to complete the parameter optimization design.Experiments were performed in a cert ain pose space.According to the objective function of parameter optimization,the integrated measurement error under different parameters was obtained,and the minimum value was 75.6% smaller than the maximum value.The simulation results show that the optimized non-coplanar target parameters can effectively improve the accuracy of monocular vision system pose measurement.The curve features that are often applied to monocular visual measurement s are studied.The circle is the most common and simplest curve feature,but a single circle can solve the two sets of results,and can not solve the complete pose,requiring additional constraints,so the following two curve characteristics were studied in this article: target features consisting of a single circle and a point outside the circle,and target features consisting of three coplanar circles.In this paper,firstly,we study the target features composed of a single circle and a point outside the circle.We take the radius of the circle and the distance from the center of the circle as the target parameters of the target,and establish a pose measurement algorithm model based on a single circle and a point outside the circle.On this basis,an objective function that optimizes the characteristic parameters of the curve is proposed.Through software programming,the pose is solved and the characteristic target parameters of the curve are optimized.In the experiment,the characteristic of the curve is composed of a single circle and a point outside the circle.Take three groups of parameters according to the change of the radius of the circle,and take three groups of parameters according to the change of the distance from the center of the circle.Experiments were performed under a certain pose space.According to the objective function of parameter optimization,the integrated measurement error under different parameters was obtained,and the minimum value was 82% smaller than the maximum value.Experiments show that the optimized curve feature target parameters can effectively improve the accuracy of monocular vision system pose measurement.Secondly,we study the characteristics of three coplanar circular targets,and we establish a pose measurement algorithm model based on the target characteristics of the target by using the radius of the circle,the distance between the three center points,and the target.On this basis,an objective function that optimizes the characteristic parameters of the curve is proposed.Through software programming,the pose is solved and the characteristic target parameters of the curve are optimized.In the experiment,the curve feature consists of 3 coplanar circles.Take 3 groups of parameters according to the change of the circle radius,and take 5 groups of parameters according to the change of the distance between the centers.Experiments were performed in a certain pose space.According to the objective function of parameter optimization,the integrated measurement error under different parameters was obtained,and the minimum value was 85% smaller than the maximum value.Experiments show that the optimized curve feature target parameters can effectively improve the accuracy of monocular vision system pose measurement.This chapter studies three linear features of the space that intersect at two points are,and uses the angles between the three space lines as the target parameters to establish a pose measurement algorithm model based on the spatial line features,which is simplified by the positional relationship constraints of the spatial lines.On the basis of the solution process,an objective function for optimizing the characteristic parameters of the straight line is proposed.Through software programming,the position and pose are solved,and the straight line characteristic target parameters are optimized.In the experiment,the straight line feature consists of 3 straight lines intersecting two points.Take 5 sets of parameters according to the angle change between 3 straight lines.Experiments were performed in a certain pose space.According to the objective function of parameter optimization,the integrated measurement error under different parameters was obtained,and the minimum value was 52% smaller than the maximum value.Experiments show that the optimized linear feature target parameters can effectively improve the accuracy of monocular vision system pose measurement.