CN111141218A

CN111141218A - Image ranging algorithm based on multiple small markers

Info

Publication number: CN111141218A
Application number: CN201911082060.1A
Authority: CN
Inventors: 韩宁生; 庞善臣; 王昭; 董玉坤; 丁桐
Original assignee: China University of Petroleum East China
Current assignee: China University of Petroleum East China
Priority date: 2019-11-07
Filing date: 2019-11-07
Publication date: 2020-05-12

Abstract

The invention provides an image ranging algorithm based on a plurality of small markers. The invention utilizes the characteristic that modern people are more convenient to obtain photos than traditional measuring equipment, determines a local scale according to the sizes of a plurality of small markers and the pixel size ratio of the small markers in the pictures, eliminates and reduces errors through a local scale algorithm and mean value processing, and finally obtains more accurate distance. The invention is tested on a desktop computer of a Linux system, and the result shows that the distance can be accurately calculated by the designed algorithm only by using the picture shot by the smart phone, and the method has the characteristics of high accuracy and high measuring speed.

Description

Image ranging algorithm based on multiple small markers

Technical Field

The invention relates to an intelligent distance measurement method, in particular to an image distance measurement algorithm based on a plurality of small markers.

Background

The measurement of a certain distance of a certain object in life, especially the distance of a non-planar regular object, is a problem often encountered in life of people, and no matter the clothing is customized, engineering distance measurement needs to be carried out by flexibly and skillfully applying various tools such as a measuring tape and the like. Such simple range finding tools often require the assistance of multiple persons, for example, a person may have difficulty measuring his or her body size. In addition, such conventional distance measuring tools often have some situations where measurement is difficult, and although more advanced distance measuring tools can be used to try to solve the difficulty, the traditional distance measuring tools all involve a large consumption of economic cost and time cost. With the continuous progress of society, the demand of people is continuously increasing, and the traditional distance measuring tool is difficult to meet the demand of people.

In recent years, with the popularization of smart phones, everyone can easily obtain a desired photo through the smart phone, and therefore methods for measuring the size of an object by using infrared rays of the smart phone are also developed, and the methods are limited to short-distance planar distance measurement, cannot meet the requirements of medium-distance and long-distance measurement and distance measurement of non-planar regular objects, require special infrared emitting devices, and are difficult to popularize on a large scale.

The effective way for solving the current situation of distance measurement is to take images through equipment (such as a mobile phone) for distance measurement. However, the problem of wide-angle distortion and perspective deformation in picture shooting and the resulting distance measurement error can be solved well by adding small markers on the object.

Disclosure of Invention

In order to solve the mapping of a positioning target ranging segment and a scale in a picture, the invention provides a ranging algorithm based on feature matching and a small marker. The invention uses the small marker as the target to be detected, solves the two problems at the same time, respectively arranges the small markers at the two ends of the ranging section, can determine the ranging section while identifying the small markers by a characteristic matching method, and can obtain a scale by the known size of the small markers and the pixel size of the detected small markers. And then, error processing is carried out through a scale and an error correction algorithm, so that the distance of the ranging section can be obtained.

The technical scheme adopted by the invention is as follows:

a ranging algorithm based on feature matching and small markers, comprising the following parts:

A. small marker icons are designed to be easily identified and easily carried (printed).

B. And detecting the picture containing the ranging segment by using a feature matching algorithm to obtain the central pixel coordinates and the respective pixel radius of each small marker.

C. And (4) carrying out correction operation by using a local scale algorithm to obtain the distance between each small marker.

D. The distance between the small markers in the ranging section is averaged, so that random errors caused by previous operations are reduced, and the real distance of the section to be measured is obtained.

The small marker icons mentioned in part a are preferably small circle icons, and have high discrimination and many feature points through printing.

In part B, the feature matching algorithm is to perform t-feature detection through the picture to position the small marker, so as to obtain the coordinate information of the small marker;

in part C, the local scale algorithm is an error correction calculation method provided by the invention for solving the problems of non-planar regular object ranging and lens distortion effect, and the idea is to construct a local scale formula;

wherein L is the pixel distance between the centers of any two small markers, x is the integral infinitesimal, K₁Scale bar of small marker with x on 0 side, K₂Is the scale calculated for the small marker on the x-L side.

In part D, the equalization processing mentioned is based on the distance measurement section except for two ends, a plurality of small markers are placed in the section, and then a plurality of distances can be obtained by selecting different small markers for local scale algorithm for a section of distance, so that a group of values of a section of distance is obtained, the average value is obtained, and the random error of the section of distance can be reduced. This has the advantage that the measurement accuracy can be improved by increasing the number of small markers to be placed.

The technical scheme provided by the invention has the beneficial effects that:

by the distance measurement method based on the small marker, a user only needs to determine the placement position of the small marker according to requirements, and then shooting is carried out through easily-obtained equipment such as a mobile phone.

Drawings

In order to more clearly explain the technical solution of the present invention, the drawings that are needed to be used in the summary of the invention will be briefly described below.

Fig. 1 is a schematic view of the present invention applied to body size measurement, in which a model is worn with a garment having small markers, which are designed in advance, and from which we can calculate a plurality of sizes of the human body.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in further detail below.

The small markers can be constructed in a preset mode or can be obtained locally, but the small markers are required to be guaranteed to have high enough discrimination and characteristics for algorithm calculation, a user needs to place the small markers at two ends of a ranging section and at any position in the ranging section according to needs, the small markers can be placed in a segmented mode again for an irregular object, then a proper angle is selected for shooting, the basic requirement is that the ranging section and each small marker are both located in a picture, the obtained picture can give a result in real time through algorithm calculation, and the user can also select and adjust the angle to shoot for multiple times to guarantee accuracy.

Through tests, the invention can be used for measuring the sizes of objects with complicated appearance characteristics such as human bodies, and experiments show that the error of the distance measurement algorithm can be within 2cm when the size is measured.

Claims

1. An image ranging algorithm based on a plurality of small markers, comprising the following parts:

A. a method of local scale determination using a plurality of small markers.

B. And (3) a local scale ranging algorithm based on small markers.

2. The method of claim 1 based on feature matching and small tagsThe ranging algorithm of the object is characterized in that in the part A, the local scale is determined by using a plurality of small markers so as to reduce errors caused by shooting. s is expressed as the size of the markers (each marker is equally large in size), s₁,s₂,s₃....s_mThe pixel size of m small markers is represented, and the local scale corresponding to each small marker is as follows: k is a radical of₁＝s/s₁,k₂＝s/₂,k₃＝s/₃...k_m＝s/s_m。

3. The local scale distance measuring algorithm based on small markers is characterized in that the local scale distance measuring algorithm in the part B is that a certain pixel coordinate located in the middle of two small markers in a known image is used for calculating the distance from the markers by using the pixel distance relation between the two small markers to obtain the scale. The method for determining the scale comprises the following steps: x/Lk₁+(L-x)/L*k₂Where L is the pixel distance of two adjacent markers, x is the pixel distance from the first marker, k₁Scale bar for the first marker, k₂Scale bar for the second marker. Solving the pixel distance x from the first marker on the image₀The actual distance is calculated by