Study On Key Techniques Of Overall Accuracy Control For Vision Measurement

Traditional visual method can’t apply to3D topography accurate measurementin large field condition. The above phenomenon is caused by the resolution ratio andfield range of image sensors, and it is also influenced by the error caused by splicing.To solve the above contradictions, key techniques of overall accuracy control forvision measurement are studied. The final purpose is to realize the precisemeasurement of the large-size object. The key link is building high precision globalcontrol net. The information of the3D morphology of the object to be tested isbounded in the net that is constituted by high positioning precision control points, thusit can solve the above contradictions. Visual control points are placed on the testedobject. Visual information of control points are taken by camera. Then the imageinformation is processed by high precision image processing, matching of controlpoints, and ray bundle adjustment. In the end high precision global control net isestablished through data fusion technique. This method can make the processing ofmeasurement simply, rapidly, flexibly, accurately. And it can satisfy the requirementsin industrial site condition.The main jobs are described as follows:1. The size, shape and coding capacity of global points have been designed.High precision image processing is used to extract the center of the controlpoints and identify code value of the coding points.2. Study on uncoding points’ precise matching. Matching algorithm based onepipolar geometry has been studied. A method using three images foraccurately matching has been employed to realize100%matching accuracy.3. Build global control net. Study a method of precision positioning, analysisthe factors of the building accuracy.