Where to put artificial marks on the measured object in close-range photogrammetry?

The images taken in digital close-range photogrammetry contain a large number of artificial landmarks, and measuring the image coordinates of these artificial landmarks is the basis of photogrammetry. Only according to the image coordinates of points with the same name and the corresponding photo posture can the three-dimensional coordinates of object points be calculated, and the purpose of matching points with the same name and three-dimensional reconstruction can be achieved. Therefore, accurate positioning of artificial landmarks is the main task of single image processing in digital close-range photogrammetry image recognition. At present, there are many kinds of artificial signs, mainly round signs, square diagonal signs and cross signs. This paper analyzes the characteristics of various artificial landmarks, their sub-pixel positioning methods and positioning accuracy, and summarizes the application scope of the above artificial landmarks. In the close-range photogrammetry of circular landmarks, circular artificial landmarks are widely used. The recognition and extraction process of circular landmark after lens imaging is simple, and the geometric characteristics of ellipse make it easy to locate. Various sub-pixel location algorithms have matured. Circular marking points are usually made of reflective materials. The inner ring of the sign is made of white reflective material and the outer ring is made of black material. Due to the reflection of retro-reflective materials, the project was supported by the National Natural Science Foundation of China. The coefficient is very high, and its reflection brightness is hundreds or even thousands of times higher than that of ordinary white signs under the irradiation of the same light source. At present, the sub-pixel positioning algorithms of circular signs mainly include elliptic least square fitting method, gray weighted centroid method and gray flat weighting method, all of which have reached the sub-pixel positioning accuracy. Central rural circle? What is the roundest landmark ellipse? What is the general equation for finding a plane ellipse by multiplication and fitting? Chat Printed in it? Are the coordinates of the center point of the ellipse? What is the formula for calculating the thousandth component of the ellipse center coordinate of each parameter of the elliptic equation? —? Shi Li? —"? In order to suppress the influence of image noise and improve the positioning accuracy, Li data generally need to be fitted twice. Gray weighted centroid method can be regarded as gray weighted centroid method. The gray centroid of is an equilateral square. Force is weight. What shape is in the actual formula? Force is the force gray-scale square weighted centroid method, that is, the gray-scale weight in the formula is taken as the square gray-scale square weighted centroid method, which further highlights the weight of the target gray-scale distribution and can obtain better positioning accuracy than the gray-scale weighted centroid method under ideal conditions. The following is the data that Lu et al. used the above three algorithms to locate the circular sign. Does the system extract image coordinates as a sign of data attack? The experimental results of the pixel table of the experimental data of circular sign positioning show that the positioning accuracy of the three methods has reached sub-pixel level. Elliptic least squares fitting method uses the gray weighted centroid method and the gray flat weighted centroid method which have the greatest influence on the center when the boundary is noisy, and uses the gray information of the whole sign, so the positioning accuracy of elliptic least squares fitting method is the lowest. Although the accuracy of the weighted centroid method based on gray average is higher than that of the weighted centroid method based on gray average, when impulse noise exists in the image, the weighted centroid algorithm based on gray square is sensitive to noise, which easily affects the accuracy of center positioning. Gray weighted centroid algorithm can suppress the noise in the image to a certain extent, and the positioning accuracy is relatively stable. Therefore, considering the positioning accuracy and stability at the same time, the gray weighted centroid method can be used in the positioning of circular landmark points. In recent years, in order to reduce the complexity of corresponding matching of traditional circular artificial marker points, many scholars began to study circular marker points based on encoders, the center of which is a circle, surrounded by a concentric segmented annular region. The figure shows the coding element III which has been extensively studied. According to the geometric and gray features of coding elements and the distribution of marking points on the surface of the measured object, the central coding elements of marking points are extracted, and then the sub-pixel positioning method of circular marking points is used to extract the central coding elements of marking points, which can effectively reduce the influence of noise and other factors and easily improve the robustness of the algorithm. Therefore, artificial landmarks based on coding elements have a good development prospect. Square corner sign is also a common artificial sign in close-range photogrammetry. The essence of locating the square corner sign is to extract the coordinates of the center pixel of the sign. At present, the sub-pixel positioning methods of square corner mainly include image gradient positioning method, gray moment edge straight line fitting positioning method and square corner mathematical model Wan Fang data image gradient positioning method. According to the image of the ideal one-dimensional edge as the blade curve, the image gradient line can be deduced.