Image center positioning optimization method

Abstract

The invention provides an image center positioning optimization method. The method comprises the following steps: after extracting a pixel-level edge and a sub-pixel-level edge of a target image according to gray features of a to-be-detected sample photo, determining a first center point coordinate of a first target image in the pixel-level edge by using a gray weighted centroid method, determining a first weighting coefficient of a first center point coordinate by combining a gray feature of the first target image, determining a second center point coordinate of a second target image in the sub-pixel-level edge by using an ellipse center fitting method, and determining a second weighting coefficient of the second center point coordinate by combining a roundness feature of the second target image; and finally, carrying out weighted summation on the first central point coordinate and the second central point coordinate to obtain a central point coordinate of the target image. According to the whole method, the accuracy and robustness of image center positioning under extreme measurement conditions such as target pollution or poor measurement angle are greatly improved, so that the later photo splicing and resolving precision is higher, and the measurement precision attenuation is smaller.

Description

technical field [0001] The invention belongs to the fields of computer vision and computer graphics, and relates to an image center positioning optimization method. Background technique [0002] Digital photogrammetry is an image-based measurement technology. After photographing the sample to be measured at each camera station to obtain a photo of the sample to be measured, the three-dimensional coordinates of the target are obtained by locating the image center point of the target in the photo of the sample to be measured. , and continue to perform subsequent measurements on this basis. Locating the center point of the target image in the photo of the sample to be tested is mainly analyzed based on the characteristics of the photo image, wherein the photo image feature mainly refers to the grayscale distribution and circularity of the target image corresponding to the target on the photo of the sample to be tested during the shooting and measurement process. [0003] Usual...

AI Technical Summary

Problems solved by technology

[0004] Under extreme measurement conditions such as target contamination or poor measurement angle, the target image may have image distortions such as abnormal gray distribution or distortion of roundness features. The image center positioning method or ellipse center fitting method) will lead to a large error in the center point positioning of the same target in all photos, which will lead to low accuracy of later photo stitching and calculation, and a large attenuation of measurement accuracy

Images