What Are the Main Types of 3D Scanning?

3D Scanning Technology

3D scanning technology can be divided into contact, non-contact, and destructive methods. It is widely used in measurement, reverse engineering, quality control, and related applications.

Contact Methods

Contact Scanning: Coordinate Measuring Machines

Coordinate measuring machines (CMMs) are efficient, precision measuring instruments developed in the 1960s. Their emergence was driven by the need for reliable measurement equipment to support the efficient machining of increasingly complex parts on automatic and CNC machine tools. Advances in electronics, computing, numerical control, and precision machining also provided the technical foundation for CMMs.

In 1960, Ferranti in the United Kingdom successfully developed the world’s first CMM. By the end of the 1960s, more than 30 companies in nearly 10 countries were manufacturing CMMs, although the products were still at an early stage. From the 1980s onward, companies including ZEISS, Leitz, DEA, LK, Mitutoyo, Ferranti, and Moore continued to introduce new products, leading to rapid CMM development. Modern CMMs can perform complex measurements under computer control, exchange information with CNC machine tools to control machining, and support reverse engineering based on measurement data.

Today, CMMs are widely used in mechanical manufacturing, automotive, electronics, aerospace, and defense industries. They have become essential equipment for modern industrial inspection and quality control.

CMM Components and Operating Principle

A CMM is a typical mechatronic system consisting mainly of mechanical and electronic systems.

1. Mechanical system: This system generally consists of three orthogonal linear-motion axes. In a typical structure, the X-axis guideway system is mounted on the worktable, the moving bridge beam forms the Y-axis guideway system, and the Z-axis guideway system is installed in the central slide. Each axis is equipped with a scale for measuring displacement. Manual handwheels and motor drives, whether powered or CNC-controlled, are generally located near the axes. The probe used to contact the surface of the inspected part is mounted at the end of the Z axis.
2. Electronic system: This system generally includes a scale-counting system, a probe-signal interface, and a computer. It acquires coordinate-point data from the measured part and processes the data.

The operating principle of a CMM is to convert measurement of geometric features into measurement of the coordinates of a series of points on those features. After point coordinates are obtained, mathematical calculations determine dimensions and geometric tolerances. For example, when measuring a cylindrical hole, several points can be probed in a plane perpendicular to the hole axis. Their spatial coordinates can then be used to calculate the hole diameter, center coordinates, roundness error, cylindricity error, and the positional relationship of the hole axis. This demonstrates the high versatility and flexibility of a CMM: in principle, it can measure any geometric feature and parameter of any workpiece.

The Complementary Role of Articulated-Arm Measuring Machines

An articulated-arm measuring machine is also a contact measuring device. Unlike a CMM, it does not use a precision worktable, column, or guideways. Instead, it uses an articulated arm with multiple degrees of freedom and can serve as a flexible coordinate measuring machine.

A sensor can be installed at the arm end. Rotary encoders record the rotation angle of each joint, and the sensor coordinates in space are calculated using kinematic principles. During use, an operator holds the measuring arm and brings the probe into contact with the inspected surface. Pressing a button records the coordinates and probe-handle direction, which are then transmitted through a serial cable to the measuring software. This type of machine is almost unrestricted by direction and can measure in any orientation within its work envelope. It is generally used for reverse-engineering measurement of large sheet-metal dies, including manual arms from CIM-CORE and FARO.

Compared with CMMs, articulated-arm measuring machines have somewhat lower accuracy but a larger measurement range and fewer limitations related to workpiece size and shape. They are compact, flexible in use, and suitable for in-process measurement. In other respects, they are similar to CMMs.

Non-Contact Methods

Non-contact measurement mainly uses optical principles to acquire data. Common methods include laser ranging, laser triangulation, structured light, image analysis, and interferometric measurement. Optical measuring equipment offers high measurement speed and a high degree of automation. Because there is no contact pressure or friction, it can avoid measurement errors caused by deformation of the part under force, making it well suited to rapid, large-scale data acquisition from complex models.

The basic principle of optical measurement is to convert a physical analog quantity into coordinate data for the part surface through appropriate algorithms.

Laser Ranging

Laser ranging calculates the distance between a measured point and a reference plane from laser-beam time of flight, the speed of light, and the atmospheric refractive index. Because direct time measurement is difficult, continuous-wave phase measurement is commonly used in phase-based rangefinders. This method can measure distances of several hundred meters with typical millimeter-level accuracy.

Laser Triangulation

Laser triangulation uses the position and angle between a light source and an imaging sensor, such as a camera, to calculate the spatial coordinates of points. It can be used for point, line, or surface measurement.

1. It has a long working distance and can measure a workpiece even when it is relatively far from the part surface.
2. It has a large measurement range. Nonlinear errors caused by a large range can be calibrated and corrected with software.
3. The probe does not contact the measured object, so it can acquire data from soft-material surfaces and measure sharp edges, depressions, and other complex contours effectively.
4. Data acquisition is fast. For large surfaces, data can be acquired quickly on a CMM or CNC machine tool without probe compensation.
5. It is relatively expensive, and stray reflections and surface features such as vertical walls can affect acquisition accuracy.
6. It is sensitive to the measured material, surface roughness, and reflectivity, and its accuracy is generally slightly lower than that of a CMM.

Structured Light and Gray-Code Technology

Structured-light methods project a specific light pattern, such as stripes or grids, onto an object surface. The three-dimensional coordinates of the surface are calculated by capturing how the reflected pattern is deformed by the curved surface. A typical approach is projected gratings, which obtain height information from deformation of grating stripes caused by changes in surface elevation.

Unlike passive image analysis, structured light requires active projection of an artificial light source. Passive stereo vision uses one or more cameras to photograph an object from different angles and calculates depth from image disparity. Binocular structured-light systems combine the advantages of both approaches and can further improve measurement efficiency.

Binocular structured-light measurement can acquire a large amount of data in one operation and is suitable for efficient measurement of complex surfaces. However, results are affected by the optical properties of the object surface, making transparent or black surfaces more difficult to measure. Accuracy is generally slightly lower than for contact measurement.

The ATOS optical scanner from GOM in Germany is a typical structured-light measuring system. It projects grating stripes onto an object surface, uses two high-resolution cameras to capture the deformed-stripe images, and calculates three-dimensional surface coordinates. It can acquire complete point-cloud data for complex structures in a very short time. The ATOS II has a single-measurement area of approximately 280 mm by 350 mm, can capture up to 1.3 million scan points, and can achieve single-measurement accuracy of plus or minus 0.03 mm. For larger objects, complete models can be created by combining measurements from multiple angles, with an assembly accuracy of approximately 0.1 mm/m.

Gray code combined with sinusoidal phase shifting is a commonly used encoding method in structured light. Gray code first divides the measurement space into regions to determine the approximate position of each point. Phase shifting is then used for precise positioning within each region.

The core characteristic of Gray code is that adjacent codes differ by only one bit. This avoids errors that can occur when multiple bits change at the same time, providing high reliability and a low decoding error rate for large-range measurement. Its main characteristics include:

1. Reliable encoding with a low error rate.
2. It is a weighted code. Different Gray codes cannot be compared directly and must first be converted to binary code for processing.
3. The number of 1s in a codeword can be used to determine whether the corresponding value is odd or even.

Multi-frequency heterodyne phase-grating technology uses a DLP projector to project computer-generated digital gratings onto an object surface. Two cameras synchronously capture the deformed stripes, and phase calculations then produce three-dimensional point-cloud data for the object surface. This approach allows flexible adjustment of grating frequency while maintaining high projection accuracy.

Image Analysis

Image analysis uses the relative position of the same point in multiple images to calculate distance through disparity and thereby obtain the point’s spatial coordinates. It is also commonly combined with other optical methods to improve measurement efficiency and applicability.

Destructive Methods

Application Boundary of Destructive Scanning

An automatic serial-sectioning scanner combines layer-by-layer material removal with layer-by-layer scanning to acquire the internal and external contour data of a model. First, the part to be measured is fully encapsulated in a dedicated resin material filled with graphite powder or pigment. Once the resin has cured, it is fixed to a milling machine and machined with a very small depth of cut to create a cross-section containing both the part and resin. The section is then moved beneath a CCD digital camera for digital sampling. Because there is a distinct boundary between the encapsulating material and the part, digital image-processing techniques such as filtering, edge extraction, texture analysis, and binarization can extract the boundary contour and obtain contour-coordinate values. Cutting and analysis are repeated until coordinate data for all model sections are obtained.

Compared with industrial computed tomography, automatic serial-sectioning scanners have lower equipment and operating costs. Because they are based on pixel extraction, they can also acquire a large amount of data in each layer. Their clear disadvantage is that they destroy the measured object. A relatively mature automatic serial-sectioning scanner on the market is the RE1000 from CGI in the United States.

Industrial Computed Tomography

Industrial computed tomography (ICT) provides a non-destructive digital tomographic method for reproducing complex internal and external structures and material forms. After a physical sample undergoes tomographic scanning, ICT produces a series of cross-sectional image slices and data. These slices and data contain complete information about the workpiece cross-section contours and internal structure.

ICT extends the successful application of medical CT into industry. It can be used for reverse modeling and quality control of small, medium, and large formed parts with complex structures, as well as for technical analysis, non-destructive diagnosis of product failures and reliability, and assembly-structure analysis.