Design of fully automatic calibration machine for reference block based on machine vision


ACTA IMEKO | www.imeko.org December 2020 | Volume 9 | Number 5 | 235 

ACTA IMEKO 
ISSN: 2221-870X 
December 2020, Volume 9, Number 5, 235 - 239 

 
 

DESIGN OF FULLY AUTOMATIC CALIBRATION MACHINE FOR 

REFERENCE BLOCK BASED ON MACHINE VISION 

 

Xiang Ren1, JinYu Lu2, Zecheng Tao3 

 
1 Metrology of He nan Institute, Zhengzhou, China, 243553773@qq.com  

2 Metrology of He nan Institute, Zhengzhou, China, jinyu_lu@126.com   

3 Innovative Testing Technology Co., Ltd., Kunshan, China, tzc@cxtest.com  

 

Abstract:  

In order to solve the problems of the existing 

standard Brinell hardness machine such as low 

automation, the time-consuming indentation 

measurement and the human influence in 

indentation measurement, a calibration machine for 

reference block was designed. It shows that this 

machine can reduce the influence of the temperature 

difference between the hardness block and the 

measurement platform; it eliminates the time for the 

indentation measuring system to find the 

indentation and juxtapose it with the centre of the 

field of view and implements fully Brinell hardness 

automatic calibration.  

Keywords: automatic indentation measuring 

system; machine vision technology 

1. INTRODUCTION 

The Brinell hardness test method is the simplest 

and fastest method in the mechanical property test 

of materials. It is one of the main methods for 

determining the mechanical properties of materials, 

inspecting product quality, determining reasonable 

heat treatment specifications and machining 

processes [1]. As a standard calibration machine 

which must have the function of reproducing and 

storing the unit value, and its reliability has special 

significance for the stable measurement 

performance which required by the standard device 

[2]. Among the existing reference block calibration 

machines, they basically adopts the loading method 

of the dead weight, which can ensure the loading 

accuracy during the measurement process, but most 

of them are produced of the 1970s.The loading 

system has the disadvantages of complicated wiring, 

large volume, single function and high failure rate, 

which brings many deficiencies and inconveniences 

to the hardness calibration work [3-5]. In the step of 

indentation measurement, the optical microscope is 

usually used, then people manually find the 

indentations and read the diameters of them, finally 

calculate the Brinell hardness value. These 

operations will probably involve personal 

measurement errors in results, and different 

operators might get different results for the same 

indentation. Moreover, measurement result may be 

different even when the same operator measures the 

same indentation multiple times. During the process, 

the operator is easy to get eye fatigue and it is more 

time-consuming to find smaller indentation points 

and the low efficiency is also the shortcoming of 

this traditional method [6][7]. 

Existing indentation measuring instruments 

include tool microscopes, image measuring 

instruments, and contact and non-contact geometry 

measuring instruments based on mechanical vision 

technology; first of all, these instruments are not 

dedicated Brinell hardness indentation measuring 

equipment, combining with the test force part of the 

machine it is difficult to realize automatic 

measurement; secondly, when measuring the 

indentation, the instruments need to move the 

reference block from the calibration machine to its 

measurement platform, this will inevitably 

introduce the influence of temperature [8]. 

Furthermore, when the diameter of the indentation 

is small, the instruments will be more time-

consuming when searching for the indentation. 

At present, a new round of revolutions, such as 

machine vision, intelligent control, new energy and 

new materials, is emerging around the world. The 

rapid development of measurement and control 

technology and machine vision technology has 

made Brinell hardness measurement surely move 

towards automation [9][10]. Based on these two 

http://www.imeko.org/
mailto:243553773@qq.com
mailto:jinyu_lu@126.com
mailto:tzc@cxtest.com


ACTA IMEKO | www.imeko.org December 2020 | Volume 9 | Number 5 | 236 

technologies and combining ISO6506-3:2014 and 

ISO6506-2:2017 developed a calibration machine 

as shown in Figure 1.  

 

Figure 1: Calibration machine for reference block 

The calibration machine is mainly composed of 

three parts: loading part, measuring part and two-

dimensional bearing platform. The loading part 

mainly includes the servo motor, reducer, lead 

screw, guide rod, measuring platform and upper and 

lower plates; the measuring part mainly consists of 

the force transducer group, laser rangefinder, 

indentation measurement system which are installed 

on the measuring platform and lead screw grating. 

2. DESIGN OF THE AUTOMATIC 

INDENTATION MEASUREMENT 

SYSTEM 

2.1. Determine the Parameters of Core 

Components 

The indentation measurement system based on 

mechanical vision is mainly composed of industrial 

cameras, telecentric lenses, and light sources. The 

selection of industrial cameras and telecentric lenses 

is mainly determined based on the measured feature 

size and calibration accuracy requirements. 

The main parameters of industrial cameras are 

resolution and sensor size; the main parameters of 

telecentric lenses are resolution, working distance, 

field of view area and magnification. 

The optimal selection of the indentation system 

is that the size of the detected feature is equal to its 

field of view area, which is enlarged by the 

telecentric lens magnification to the size of the 

target surface of the industrial camera. In this case, 

the camera pixels will be fully used to characterize 

the measured indentation, thereby reducing the cost 

of the calibration device. According to the 

resolution calculation formula (1) of the industrial 

camera, the requirements of the detected feature and 

the measurement accuracy must be obtained.  

𝑅 =
𝐿MAX
𝐿MIN

× 𝑝 (1) 

with 𝑅 – resolution; 

𝐿MAX - detected feature; 

𝐿MIN – accuracy; 

p - It indicates that several pixels are used to 

characterize the measurement accuracy. (If the 

entire image is regarded as a periodic signal with a 

minimum feature size, according to Nyquist 

sampling law, the signal must be sampled at least 2 

points per cycle to fully reproduce the signal.) 

The measured feature size is determined 

according to ISO6506-1:2014 7.4, The test force 

should be chosen so that the diameter of the 

indentation, d, lies between the values 0.24 D and 

0.6 D [11],  as there are four types of hard metal ball 

diameters: 10 mm, 5 mm, 2.5 mm, and 1 mm, the 

corresponding indentation diameter ranges are 

shown in Table 1. ISO6506-3:2014 4.6 shows the 

requirements for the measurement accuracy of the 

indentation measurement system in Table 1, [12]. 

Combine the indentation diameter range with the 

corresponding accuracy requirements to calculate 

the requirements for the resolution of industrial 

cameras.  

Table 1: Indentation range and performance requirements 

of indentation measurement system  

Hard metal 

ball diameter 

(mm) 

Indentation 

Range 

(mm) 

Diameter of 

indentation 

(mm)  

Performance 

 (mm) 

10 2.4 ~6  
d ≥ 2.5 ± 0.0020 

5 1.2 ~3  

1 ≤ d ＜ 2.5 ± 0.0010 

2.5 0.6 ~1.5  

d ＜ 1 ± 0.0005 
1 0.24 ~0.6  

2.2. Configuration of Automatic Indentation 

Measure System  

Because the range of indentation diameter is 

wide and the measure performance of different 

ranges is also different, if using a fixed-

magnification telecentric lens with an ultra-high-

pixel industrial camera can realize the automatic 

measurement of various indentations. But in this 

Industry Camera, 

Electric linear 

zoom lens and 

Lights 

http://www.imeko.org/


ACTA IMEKO | www.imeko.org December 2020 | Volume 9 | Number 5 | 237 

configuration the indentation measurement system 

needs to ensure the complete shoot of the full 

maximum indentation diameter (6 mm), at the same 

time, it must also meet the measure accuracy 

requirement (0.0005 mm) of the minimum 

indentation diameter. When 2 pixels are used to 

characterize the measurement accuracy, according 

to the resolution calculation formula, R is equal to 

24000, the camera resolution would be 

24,000 × 24,000 about 576 million resolution. The 

camera cost will be extremely high. If one camera is 

used with multiple telecentric lenses, the hardware 

configuration cannot achieve automatic 

measurement of the full diameter range.  

The use of the electric linear zoom lens will 

reduce the camera cost and achieve automatic 

measurement of the full diameter range. Firstly, the 

lens can be automatically adjusted to the specified 

magnification. When the lens enlarges a small 

indentation, the field of view will be reduced as well, 

so that the pixels can fully characterize the 

indentation.  

The automatic indentation measuring system 

uses an electric linear zoom lens with built-in 

coaxial blue light and its lens field of view can be 

adjusted to 6 mm, 2.5 mm and 1 mm, so each field 

of view can ensure the largest indentation can be 

sampled. At the same measurement accuracy 

condition, the larger indentation feature requires the 

higher resolution of the camera, so the camera 

resolution must meet the maximum indentation 

measuring requirement. When the indentation 

diameter range is between 2.5 mm and 6 mm, the 

indentation measurement accuracy is required to be 

0.002 mm. When 2 pixels are used to characterize 

the measurement accuracy, calculated as follows 

according to the resolution formula (1):  

𝑅 =
𝐿MAX
𝐿MIN

× 𝑝 

𝑅 =
6𝑚𝑚

0.002𝑚𝑚
× 2 

𝑅 = 6000 . 

 

At this time, the maximum feature of the 

indentation is equal to the field of view. Based on 

the above conditions, the pixel requirement in one 

direction of the camera is calculated to be 6000, 

since the detected indentation is a diameter, the 

number of pixels required in the other direction is 

also 6000, then ideally, the resolution of 36 million 

can meet the requirements of measurement accuracy. 

However, there are few industrial camera sensor on 

the market, generally rectangular with an aspect 

ratio of 1.33, such as 12.8 × 9.6, 6.4 × 4.8, etc. 

Therefore, in order to facilitate camera selection, 

when the number of vertical pixels is 6000, the 

number of horizontal pixels is 6000 × 1.33 is about 

8000, then the actual camera resolution selection is 

about 50 million. According to the above 

calculation process combined with the three field of 

view diameter and measurement accuracy 

requirements to calculate the corresponding 

resolution requirements, as shown in Table 2. 

Table 2: Requirement of camera resolution 

Diameter of 

indentation 

(mm) 

Performance 

(mm) 

Required 

camera 

resolution 

d ≥ 2.5 ± 0.0020 6000 

1 ≤ d ＜ 2.5 ± 0.0010 5000 

d ＜ 1 ± 0.0005 4000 

 

According to Table 2 it can be found that when 

the measurement accuracy of the indentation is 

0.002 mm and the field of view area is 6 mm, the 

requirement of the camera resolution is the highest. 

It is concluded that the camera with about 50 million 

pixels and equipped with electric linear zoom lens 

can meet the automatically measurement and 

accuracy requirements.  

3. DESIGN THE FORCE PART OF THE 

CALIBRATION MACHINE 

The test force of this calibration machine is 

provided by high-precision force transducers, 

according to the ISO 6506-1: 2014-7.3, and the 

requirement in ISO 6506-3: 2014-4.4, selects ISO 

376 class 0.03 high-precision force transducers and 

range of 30 kN, 3 kN and 300 N to be installed on 

the measurement platform. When the test condition 

is selected, the corresponding force transducer will 

move to the loading position, this process is 

implemented by the linear guide rail, slider, screw 

rod, grating and servo motor on the measurement 

platform. 

When the force transducer at the loading position, 

the axis of the force transducer coincides with the 

axis of the loading screw to avoid the generation of 

lateral force, and the loading and unloading of the 

test force is then completed through the 

measurement and control system and the loading 

mechanism. 

The two-dimensional bearing platform is 

composed of steel plate, screw, grating ruler, linear 

guide, slider and servo motor. The platform is 

precisely controlled by software to move the 

reference block. And write the requirements of 

http://www.imeko.org/


ACTA IMEKO | www.imeko.org December 2020 | Volume 9 | Number 5 | 238 

ISO6506-3:2014 7.8 indentation selection into the 

software, so as to improve the efficiency of loading 

point selection and make better use of the test area 

of the reference block. 

 

 

Figure 2: Main parts for moving the force transducer and 

two-dimensional bearing platform 

4. AUTOMATIC CALIBRATION 

PROCEDURE 

4.1. The Key Point of Keeping Indentation 

Measurement Accuracy During The 

Calibration Procedure 

The principle of machine vision image 

measurement is to map the length in the world 

coordinate system to the pixels in the camera 

coordinate system in combination with the imaging 

system model. The better reproduction of the 

calibration conditions is most conducive to ensuring 

the accuracy of machine vision indentation 

measurement. Before the indentation measure 

system calibration, the camera should be installed 

vertically on the measurement platform to reduce 

the error caused by perspective distortion. The rest 

influence factors such as lens distortion and 

tangential distortion are caused by the hardware 

itself. Once the system calibration is completed, 

they are not easy to change during measurement. 

The calibration process is to map the real length to 

the indentation measurement system pixel and find 

its work distance. The work distance of the 

indentation test system is the most important 

condition to ensure accuracy, because the change of 

the distance will directly affect the mapping 

relationship between pixel points and world 

coordinates. In order to ensure the reproducibility of 

the work distance, the laser rangefinder, the grating 

of monitoring the lead screw combined with 

advanced calibration procedures to ensure this 

condition. 

4.2. Reference Block Calibration Procedure 

The first step is to place the reference block on 

the two-dimensional bearing platform, and then 

select the test force and light source according to the 

test condition. 

The second step is to determine the longitudinal 

sample point of the indentation measurement 

system, due to the inconsistency of the reference 

block thickness and it has the surface problem of 

straightness and parallelism, so the indentation 

measurement system must also move longitudinally 

to ensure the reproducibility of the calibration 

distance. To solve the above problem, the hardware 

configuration and calibration methods are as 

follows:  

a. The servo motor on the measuring platform 

moves the laser rangefinder to the horizontal 

working position through the linear guide, slider, 

screw and grating. In the working position, the 

optical point of the laser rangefinder coincides with 

the axis of the lead screw. 

b. Then the laser rangefinder combined the 

thickness of the reference block preloading point 

with work distance of indentation measurement 

system find the longitudinal sample point, and the 

grating which monitors the lead screw records this 

longitudinal position. This position will be the 

longitudinal sample position of indentation 

measurement system, and then the measurement 

platform rises back to the initial position. 

The third step, the force transducer 

corresponding to the test force is moved to the 

horizontal working position. In the working position, 

the axis of the force transducer coincides with the 

axis of the lead screw to avoid the generation of 

lateral force, and then the test force is applied 

through the measurement and control system and 

the loading mechanism. After completing the above 

operations, the measurement platform rises back to 

the initial position.  

The fourth step: First, the indentation 

measurement system is moved to the horizontal 

working position. At this position, the centre of the 

lens coincides with the axis of the lead screw, and 

then the software combine the lead screw with its 

grating to move the indentation measurement 

system to the longitudinal sample position recorded 

in the second step b.  

In the fifth step, the software processes the 

sampled image and calculates the hardness value. 

After the work is completed, the measurement 

Linear guide rail 

Slider 

Grating 

Measurement 

Platform 

Guide rod 

Screw 

Linear guide rail 
Slider 

Servo motor 

Screw 

Reference block 

Linear guide rail 
Slider 

http://www.imeko.org/


ACTA IMEKO | www.imeko.org December 2020 | Volume 9 | Number 5 | 239 

platform and each measurement component are 

restored to the initial position, and then the next test 

point is automatically or manually selected 

according to the requirements by the two-

dimensional platform. 

5. SUMMARY 

This calibration machine has the following 

advantages: first, it completes the automatic 

calibration of the Brinell standard block through an 

advanced calibration program combined with the 

measurement and control system and the automatic 

indentation measurement system. 

Second, due to the application of test force and 

the indentation are both on the two-dimensional 

pressure platform, this reduces the influence of the 

temperature by placing a certain number of 

reference blocks on the two-dimensional bearing 

platform in advance. 

Third, according to the calibration procedure, the 

laser rangefinder, force transducer, and indentation 

measurement system will be moved to the same 

horizontal working position by the measurement 

platform during work, which ensures that the 

indentation is located in the centre of the field of 

view when sampling indentation, thus saving the 

time of finding the indentation and juxtapose it with 

the centre of the field of view so this improves the 

calibration efficiency.  

When other geometry measuring instruments 

based on mechanical vision technology measured 

the reference block, the indentation is formed. 

Therefore, it can only use the surface other than the 

indentation to determine the working distance of the 

instrument and the reference block also has 

problems with straightness, parallelism and surface 

roughness, so it will affect the measurement results 

to a certain extent. Compared to the instruments, 

this calibration machine’s work distance is obtained 

before the indentation formed. Therefore, the 

influence of the working distance on the 

measurement result can be reduced to a certain 

extent. 

 

 

6. REFERENCES 

[1] Yongchang Huang, Cheng Zhou, Tao Chen, Song 

Li, Guiping Qiu “Development of Brinell Hardness 

Indentation Automatic Measuring Device Based on 

High Definition Digital Image Analysis”, 

Measurement Technique, pp. 10-14, July 2019. 

[2] Wei Zhang, Xin Luo, Jiachun Lin, “Automatic 

control system for standard Brinell hardness 

machine based on PLC”, Journal of Mechanical & 

Electrical Engineering, pp. 211-212, Feb 2020. 

[3] Jizeng Tao, Wei Shi, Jinying Li, Zhongyan Wu, 

“Application of Piezoelectric Ceramics to the Force 

Loading System of the Brinell Hardness Standard 

Machine”, Measurement & Control Technology, 

pp. 72-73, Nov 2011. 

[4] Yanan Yu, Wei Shi, Wu Zeng, Jizeng Tao, Ming 

Xu, Jinying Li, “Development of electric Brinell 

hardness standard device”, China Metrology, pp. 

72-74, Sep 2010. 

[5] Jintao Le, “Development Status and Direction for 

the Hardness Testing Machine Manufacturing 

Industry in China”, Physical Testing and Chemical 

Analysis (Part A: Physical Testing), pp. 379-381, 

June 2017. 

[6] Jizeng Tao, Wei Shi, Jinying Li, “The Brinell 

Hardness Measurement Device Based on CCD and 

Grating Ruler”, Metrology & Measurement 

Technology, pp. 18-20, June 2011. 

[7] Wutingting Di, “Computer Measurement System of 

Brinell Hardness Indentation ”, Digital Technology 

and Application, pp. 48+51, Nov 2016. 

[8] Manlong Chen, “The fast calibration method for 

image measuring instrument”, Modern 

Manufacturing Engineering, pp. 126-130+140, 

Aug 2017. 

[9] Weiguang Lu, Guoyuan Wang，HaiLiang Wang, 

Zheng Cai “A Brief Talk on Geometric 

Measurement Technology and Its Application in 

Industrial Manufacturing Industry”, Aviation 

Precision Manufacturing Technology, pp. 43-46, 

May 2019. 

[10] Yongchang Huang, Wei Huang, Bin Chen, Yu Bai, 

Mingtao Feng, Feiyu Zhao, “Application of Brinell 

Hardness Indentation Automatic Measuring Device 

in Verification”, Metrology & Measurement 

Technology, pp. 114-117, Jan 2016. 

[11] ISO 6506-1:Metallic materials –Brinell hardness 

test–Part 1: Test method[s]. Switzerland Gevena: 

International Standardization Organization,2014 

[12] ISO 6506-3:Metallic materials –Brinell hardness 

test–Part 3: Calibration of reference blocks[s]. 

Switzerland Gevena: International Standardization 

Organization,2014 

 

 

 

http://www.imeko.org/