New close-range photogrammetry method based on grain-lacking object

摘要

Close-range photogrammetry is a significant method that can detect size, shape and position of objects for its conveniences and high accuracy. But in some extreme environment, the conventional method is difficult to match the request of measurement for there are still many measurement work can not complete using traditional method. This paper has development a new method to measure the section of objects using the single camera measurement model. In order to achieve the purpose, there are three main parts in this paper. Firstly, two extraction method of laser fringe is presented, their extraction precision and time is compared via extracting laser fringe from images with different Gauss noise. Steger method's precision is higher than curve fitting method. But curve fitting method cost less time than Steger method. Secondly, we have improved the traditional Autobar to adapt the dark measure environment. Considering retro-reflective targets and common black-white targets can not be recognized easily while without strobe light or lack of illumination, the retro-reflective material of traditional Autobar is replaced with LED light to be recognized easily in image without strong flicker when photographing. At last, a simulation experiment is taken to demonstrate the whole measurement process and validate the new single camera measurement model' feasibility. The final results of simulation experiments showed that the newly presented measurement model has its feasibility. This measurement model greatly improves the measurement efficiency and makes the measurement work more flexible.