Irregular object volume measurement method based on compressed sensing

Abstract

Provided is an irregular object volume measurement method based on compressed sensing. Firstly, a system is calibrated, and a pixel equivalent of the system is obtained; secondly, original images of an object to be measured in the top view direction, the left view direction and right view direction are obtained respectively by using three cameras, coordinate parameters of the maximum peripheral length image of the object to be measured are obtained through the top view image, reconstruction ranges corresponding to the left view image and the right view image are segmented according to top view maximum length pixel count, three-dimensional reconstruction is carried out on the left view image and the right view image respectively through a light and shade shape restoring method, a part of coordinate points obtained through the light and shade shape restoring method are selected randomly, the part of the coordinate points are regarded as measurement values, a compression measurement matrix and a sparse base are selected, the original coordinate of the left view image and the original coordinate of the right view image are reconstructed, the accuracy of coordinate values is improved, the volume of a left view image reconstruction object and the volume of a right view image reconstruction object are calculated respectively under image coordinates, and then summation is carried out; lastly, the actual volume of the object to be measured in a world coordinate system is calculated according to the pixel equivalent. The volumes of irregular objects can be measured through the irregular object volume measurement method based on the compressed sensing.

Description

technical field [0001] The invention relates to a method for measuring the volume of irregular objects by using digital image processing technology, in particular to a method for measuring the volume of irregular objects based on compressed sensing. Background technique [0002] In real life and industrial fields, some irregular materials need to measure their volume, such as the online detection system of industrial raw materials. With the development of measurement technology, the measurement environment of irregular objects is complex and changeable, and the measurement methods are constantly innovating. [0003] The traditional measurement method is to use a tape measure to measure the geometric shape of a regular object on the spot. Although it is simple and easy to implement, it needs to sort out the observed objects. Manual on-site measurement has poor timeliness and low accuracy. It is not suitable for the measurement of large objects, and it is also difficult to ach...

AI Technical Summary

Problems solved by technology

Binocular (multi-) visual measurement is to obtain object image information from different angles. The relative positions of multiple cameras must be fixed accurately. At the same time, it is necessary to ensure the synchronization and stability of multiple cameras in shooting, which is difficult to achieve in control. Basically, only a small range of applications can be realized, and feature points for matching between images need to be found, and feature point matching is a difficult point in image processing. Therefore, the volume calculation operation after 3D reconstruction using multiple images is complex and computationally intensive. Large, not suitable for problems such as dynamic scenes

There is no matching of image feature points in the monocular vision measurement method. It directly uses the known geometric features in the scene to obtain the three-dimensional structure of the object under the meaning of an affine transformation. The requirements for equipment are simple and no need for multiple images. Matching, low cost, easy to implement, saves time and improves efficiency. Shading Restoration Shape (SFS) method is the most commonly used monocular vision method. height, but due to the low reconstruction accuracy of the assumed conditions, the existing methods for measuring the volume of irregular objects limit their application in industrial processes to a certain extent

Images