CN111207691A

CN111207691A - Method for realizing measurement of optical fiber shape

Info

Publication number: CN111207691A
Application number: CN202010103172.7A
Authority: CN
Inventors: 秦晓飞; 徐敏敏; 赵颖; 蔡锐; 刘燕; 张学典
Original assignee: University of Shanghai for Science and Technology
Current assignee: University of Shanghai for Science and Technology
Priority date: 2020-02-19
Filing date: 2020-02-19
Publication date: 2020-05-29

Abstract

The invention relates to a method for measuring the shape of an optical fiber, wherein different stereoscopic vision positioning labels are arranged on the surface of the optical fiber at equal intervals; acquiring internal and external parameters of the camera and the position relation of a plurality of cameras by using a camera calibration method; acquiring images of the visual positioning labels by using a multi-view camera, and sending the acquired images to a computer system to obtain the position and the posture of each visual positioning label through calculation so as to obtain the shape of the optical fiber; using a given light source as an input light input optical fiber, and obtaining output light of the optical fiber through measurement; and fitting the corresponding relation between the input light and the output light of the optical fiber and the shape of the optical fiber by using a deep learning method, thereby obtaining a shape sensing model of the optical fiber under the conditions of the input light and the output light. The method can directly obtain the shape of the optical fiber under the condition of known input light and output light.

Description

Method for realizing measurement of optical fiber shape

Technical Field

The invention relates to a detection technology, in particular to a method for realizing measurement of an optical fiber shape.

Background

An optical fiber is a short term optical fiber, and is an optical element that propagates light along a curved path by using the principle of total reflection. The optical fiber is composed of a core and a cladding, wherein the core is made of glass or plastic, the cladding is also made of glass or plastic, and a plastic protective outer sleeve is arranged outside the cladding. Light is generally transmitted in the optical fiber, the light guiding capability of the optical fiber depends on the refractive indexes of the fiber core and the cladding, and the refractive index of the cladding is usually slightly smaller than that of the fiber core, so that the optical signal can be transmitted in the fiber core in a closed mode, and the fiber core is protected. The fine fibers are enclosed in a sheath so that they can be bent without breaking. By utilizing the advantages of high reliability, low cost and the like of the optical fiber, the optical fiber is generally used in the communication aspect before, the signal to be transmitted is modulated into the optical signal to be transmitted along the optical fiber, and the optical signal is demodulated at a receiving end to obtain the original signal. Because of the characteristics of the optical fiber, the application of the optical fiber is more and more extensive at present, the optical fiber is used for transmitting signals and utilizing the characteristic of being capable of bending, and the optical fiber can be embedded into a flexible surgical instrument and used for a surgical navigation system. And is therefore also important for the shape of the fiber when measured.

In the field of visual positioning, binocular or multi-view cameras are commonly used. The binocular stereo vision is a method for acquiring three-dimensional geometric information of an object by acquiring two images of the object to be measured from different angles by using imaging equipment based on a parallax principle and calculating position deviation of corresponding points of the images. Or a monocular camera can be used for obtaining two images of the measured object from different angles, and the three-dimensional geometric information of the measured object is calculated based on the parallax principle. The method comprises the steps of obtaining the three-dimensional coordinate of a certain point in a three-dimensional space, needing to exist corresponding points of the point on the image surfaces of a left camera and a right camera, finding out the corresponding points of the point in the space on the image surfaces of the left camera and the right camera, and obtaining internal and external parameters of the cameras through camera calibration so as to determine the three-dimensional coordinate of the point. Sometimes one camera is blocked for other reasons, and multiple cameras are also used to improve the accuracy of the measurement.

Disclosure of Invention

The invention provides a method for measuring the shape of an optical fiber, which aims at the problem of optical fiber application, and the method comprises the steps of measuring the shape of the optical fiber by using an optical fiber sensitive to the shape and utilizing a multi-view visual positioning method, measuring input light and output light of the optical fiber, fitting the input light and the output light serving as known information with the shape of the optical fiber to obtain a corresponding relation, and thus directly obtaining the shape of the optical fiber under the conditions of known input light and output light.

The technical scheme of the invention is as follows: a method for realizing optical fiber shape measurement comprises mounting a plurality of different visual positioning labels with equal intervals on the surface of an optical fiber to be measured, and labeling the visual positioning labels; the multi-view cameras are orderly arranged around the optical fiber to be detected, each multi-view camera obtains images of the optical fiber to be detected from different angles and sends the images to the computer system, and the computer system obtains three-dimensional geometric information of the object by calculating the position deviation of corresponding points; sending the given optical fiber input light and the optical fiber output light obtained by the output light measuring device to a computer system; and fitting by a computer system to obtain the corresponding relation between the input light and the output light and the shape of the optical fiber, changing the shape of the optical fiber to be detected for multiple times, and repeating the steps to obtain the shape sensing model of the optical fiber to be detected under the conditions of the known input light and the known output light.

The method for realizing the measurement of the shape of the optical fiber comprises the steps of changing the shape of the optical fiber for multiple times, measuring the shape of the optical fiber by using a multi-view visual positioning method, measuring input light and output light of the optical fiber for multiple times, fitting the corresponding relation between the input light and the output light and the shape of the optical fiber by using a depth learning method, and thus obtaining a shape sensing model of the optical fiber under the conditions of the input light and the output light.

The invention has the beneficial effects that: the method for realizing the measurement of the shape of the optical fiber has the advantages of high precision, high efficiency, high sensitivity and the like, and the shape of the optical fiber can be directly obtained through a fitted model under the known conditions of input light and output light.

Drawings

FIG. 1 is a schematic diagram of an implementation of the method for measuring the shape of an optical fiber according to the present invention;

FIG. 2 is a diagram of a fiber optic label in the method of the present invention;

FIG. 3 is a block diagram of the working flow of the method for measuring the shape of an optical fiber according to the present invention.

Detailed Description

The method for measuring the shape of the optical fiber obtains the shape of the optical fiber by a multi-view visual positioning method, and fits the corresponding relation between input light and output light and the shape of the optical fiber, so that the shape of the optical fiber is directly obtained according to the corresponding relation under the known conditions of the input light and the output light.

Fig. 1 is a schematic structural diagram of an implementation method for measuring the shape of an optical fiber, which includes a shape-sensitive optical fiber 101, a multi-view camera 301, a visual positioning tag 201, input light 103, an output light measuring device 105, and a computer system 106. As shown in fig. 2, a multi-view camera 301 captures image information of a plurality of different visual positioning tags 201 on the surface of the optical fiber 101, and directly sends the acquired images of the plurality of different visual positioning tags 201 to the computer system 106. Input light 103 for a given fiber 101 and output light from an output light measurement device 105 are sent to a computer system 106. The computer system 106 calculates the shape of the optical fiber 101 from the image information, and fits the correspondence between the input light 103 and the output light and the shape of the optical fiber 101 by using a deep learning method. Wherein the optical fiber 101 shown in fig. 1 is a shape-sensitive optical fiber 101; the input light 103 uses a specific light source; the multiple multi-view cameras 301 are orderly arranged around the optical fiber 101, each multi-view camera 301 can acquire images of an object to be measured from different angles by using imaging equipment based on a parallax principle, and three-dimensional geometric information of the object is acquired by calculating position deviation of corresponding points, so that the accuracy of the system can be improved, only four multi-view cameras 301 are drawn on the image, and the rest are represented by ellipses.

Fig. 3 is a block diagram of the working flow of the method for measuring the shape of the optical fiber according to the present invention. The specific work flow of the block diagram is as follows: first, the multi-view camera 301 and the output light measuring device 105 are initialized, the multi-view camera 301 is connected to the computer system 106, and the output light measuring device 105 is also connected to the computer system 106. Then, the optical fiber 101 is placed in the test area, the data of the multi-view camera 301 is read, then, each visual positioning label 201 is identified, the computer system 106 calculates the position deviation of the corresponding point on the image obtained by each multi-view camera 301 by utilizing the parallax principle, and calculates the three-dimensional coordinates, the pitch angle, the yaw angle and the roll angle of a plurality of different visual positioning labels 201 installed on the surface of the optical fiber 101 in the real space, thereby calculating the position and the posture of the visual positioning label 201 and obtaining the shape of the optical fiber 101. And reading data of the output light measuring equipment 105, fitting to obtain a corresponding relation between the shapes of the input light 103 and the output light and the optical fiber 101, changing the shape of the optical fiber 101 for multiple times, and continuing the process to obtain a shape sensing model of the optical fiber 101 under the condition of knowing the input light 103 and the output light.

The visual alignment tag 201 may be mounted on the optical fiber 101 at equal intervals of 5cm or less as shown in fig. 2, and the visual alignment tag 201 is labeled, so that even if the optical fiber 101 is bent or knotted, the computer system 106 can obtain an accurate position and posture thereof through the label, and then obtain a more accurate shape of the optical fiber 101. In order to obtain a higher accuracy of the shape of the optical fiber 101 measured by the present invention, the pitch of the visual alignment marks 201 mounted on the surface of the optical fiber 101 is as small as possible.

The multi-view visual positioning method comprises the following steps: the method comprises the steps of using optical fibers sensitive to shapes, installing different stereoscopic vision positioning labels on the surfaces of the optical fibers at equal intervals, obtaining internal and external parameters of a camera and the position relation between a plurality of cameras by a camera calibration method, capturing images of the visual positioning labels through a multi-camera, sending the captured images to a computer system, and obtaining the position and the posture of each visual positioning label through computer statistics, so that the shape of the optical fibers is obtained. The shape of the optical fiber is changed for many times, and the shape of the optical fiber is measured by using a multi-view visual positioning method.

The shape sensing model is characterized in that the shape of the optical fiber is changed for multiple times, the shape of the optical fiber is measured by using a multi-view visual positioning method, the input light and the output light of the optical fiber are measured for multiple times, and the corresponding relation between the input light and the output light and the shape of the optical fiber is fitted by using a deep learning method, so that the shape sensing model of the optical fiber under the conditions of the input light and the output light is obtained. In practice, the shape of the optical fiber can be obtained directly using the established shape sensing model under known input and output light conditions.

By the method, the relation between the shape of the optical fiber and the input light and the output light is fitted, and the optical fiber can be embedded into a flexible surgical instrument for shape measurement and is used for a surgical navigation system.

Claims

1. An implementation method for measuring the shape of an optical fiber is characterized in that a plurality of different visual positioning labels with equal intervals are arranged on the surface of the optical fiber to be measured, and the plurality of visual positioning labels are labeled; the multi-view cameras are orderly arranged around the optical fiber to be detected, each multi-view camera obtains images of the optical fiber to be detected from different angles and sends the images to the computer system, and the computer system obtains three-dimensional geometric information of the object by calculating the position deviation of corresponding points; sending the given optical fiber input light and the optical fiber output light obtained by the output light measuring device to a computer system; and fitting by a computer system to obtain the corresponding relation between the input light and the output light and the shape of the optical fiber, changing the shape of the optical fiber to be detected for multiple times, and repeating the steps to obtain the shape sensing model of the optical fiber to be detected under the conditions of the known input light and the known output light.

2. The method of claim 1, wherein the shape of the optical fiber is changed many times, the shape of the optical fiber is measured using a multi-view visual positioning method, the input light and the output light of the optical fiber are measured many times, a deep learning method is used to fit the corresponding relationship between the input light and the output light and the shape of the optical fiber, so as to obtain a shape sensing model of the optical fiber under the input light and the output light, and when the shape measurement is actually performed, the shape of the optical fiber can be directly obtained by using the established shape sensing model under the known input light and output light conditions.