In recent years, the development of 3D data measurement technology and the gradual popularization of 

related coordinate measuring equipment make the acquisition of 3D data more and more convenient. 3D 

data measurement technology is an important part of 3D reconstruction technology. Currently, common 3D

data measurement methods are divided into contact and non-contact types [1-4].

The contact method is to use some instruments to directly measure the three-dimensional data of the 

scene through actual contact. The typical representative is the coordinate measuring machine (CMM), 

which is used to measure some small volume important parts in the field of industrial measurement [5]. In 

the manufacturing process, using a CMM to measure and obtain dimensional data for mechanical parts 

and molds is one of the most effective methods, because it can replace a variety of surface measurement 

tools and expensive combination gauges, and reduce the time required for complex measurement tasks 

from hours to minutes, which is not achieved by other instruments. The data obtained by the contact 

method has high accuracy and is less affected by environmental factors, but the cost is high, and it is only

 suitable for objects with more standard surface geometry. The application of the object to be measured is

 limited to some extent in cases where the surface of the object to be measured is complex, the surface 

of the object is soft and easy to deform, and the surface geometry of the object is less. What's more, it is very difficult to obtain large amounts of three-dimensional point cloud data of objects using contact 

measurement methods.

  Different from the contact method, the non-contact method does not touch the measured object during 

measurement, maintains a certain distance from it, and obtains target data through light, sound, magneticfield and other media [6]. Although the measurement accuracy is lower than that of the contact 

measurement method, with the development of science and technology, the measurement accuracy of the 

non-contact method has been greatly improved. In the field of 3D reconstruction, the non-contact 

measurement method is more widely used than the contact measurement method.

Non-contact scanning is divided into active scanning and passive scanning according to different scanning

methods. As shown in Figure 2, active scanning commonly includes 3D laser scanning technology [7], 

Time-of-Flight camera [8], structured light system [9] and so on.

  Three-dimensional laser scanning technology is also known as reality reproduction technology, laser 

points or linear laser beams are issued for the object or scene to be measured, and the light reflected by the object to be measured is received through the photoelectric element, and the timer determines the time of the laser beam from transmission to reception, so as to calculate the distance between the object 

to be measured and the device. This method can obtain more accurate depth information, and can record 

a large number of dense point coordinates, the disadvantage is that the price is high and the data 

information is large. Omari et al. [10] proposed to apply 3D laser scanning technology to the generation of 

progress reports in construction projects. Data generated by scanning construction sites with 3D laser 

scanners at different times were then used to estimate the workload performed within the time interval of 

two consecutive scans, so as to improve the speed and accuracy of data collection from construction 

sites. Ghahremani et al. [11] proposed a method to obtain point clouds with a handheld 3D laser scanner to

 ensure the quality of high-frequency mechanical impact processing. The method carried out different 

degrees of impact processing on five welding patterns, then used a handheld 3D laser scanner to scan the

 point cloud data of the welding joint, and determined different impact processing by measuring the 

geometric parameters of the point cloud.

  The TOF camera uses modulated near-infrared light to detect the target, and the sensor calculates the 

time required for the round trip of the light to convert the distance between the object and the device to 

obtain the depth information. Combined with the traditional camera shooting, the three-dimensional outline

 of the object can be represented by different colors representing different distances. This method can 

quickly and easily obtain the depth data, without the need for redundant auxiliary scanning equipment, the 

disadvantage is that the price is still high compared with the ordinary camera, the measurement results are

 affected by the properties of the measured object, and the system error and random error have a 

significant impact on the results. Heide et al. [12] changed the scene reconstruction formula into a linear 

inverse problem on the transient response of the scene, recovered objects outside the line of sight from 

second-order diffuse reflection, and used TOF cameras to obtain scene information, thus effectively 

reconstructing the scene of a large room. Koutsoudis et al. [13] used TOF cameras to carry out 3D 

reconstruction of Ottoman Empire monuments in Kessanci region, Greece, and questioned the applicability

 of the moving-Structure-Intensive Multi-view 3D reconstruction (SVM-DMVR) method through digital 

analysis of the complete 3D model of the monuments.

Structural light is a set of system structure composed of projectors and cameras. After projectors project 

specific light information to the surface of the object and after the background, the camera collects it. 

The position and depth of the object can be calculated according to the changes in the light signal caused

 by the object, and then the whole three-dimensional space can be restored [14]. The advantage of 

structured light system is that the quality of image data is better, but it requires multiple cameras to 

coordinate and work together, and it is greatly affected by lighting conditions, which has high requirements

 for the system. Deetjen et al. [15] proposed a new automatic calibration method for structured light 

systems such as camera-projector, which realizes calibration through automatic image processing and 

optimization of bunched adjustment parameters, and can be applied to quality control, autonomous 

systems, engineering measurement, motion capture and other fields. Heist et al. [16] proposed a "5D 

sensor" that quickly and accurately measures surface shape and spectral characteristics through 

structured light. The sensor provides excellent spatial and spectral resolution, excellent depth accuracy 

and high frame rate, which can be applied in fields such as medical imaging, precision agriculture or 

optical sorting. Table 1 compares the common scanning methods.