400 is seen to be expanding as it moves away from the LLP 200.

FIG. 5 shows a schematic illustration of elements of an LLP 500, including a projector 520 and a camera 540. FIG. 5 is a schematic illustration of the LLP 200 when viewed from the top with the LLP 500 looking toward object surfaces 510A, 510B. Because of the change in viewpoint, the camera 220 is to the left of the projector 210 in FIG. 4, while the equivalent camera 540 is to the right of the projector 520 in FIG. 4 in the changed viewpoint. The projector 520 includes a source pattern of light 521 and a projector lens 522. The projector lens 522 includes a projector perspective center and a projector optical axis that passes through the projector perspective center. In the exemplary system of FIG. 5, a central ray 524 of the beam of light coincides with the projector optical axis. The camera 540 includes a camera lens 534 and a photosensitive array 2641. The camera lens 534 has a camera lens optical axis 536 that passes through a camera lens perspective center 537. In the exemplary LLP 500, the camera lens optical axis 536 and the projector optical axis are both perpendicular to a plane that encompasses the line of light 523 projected by the source pattern of light 521. In other words, the plane that encompasses all the lines of light 523 is in the direction perpendicular to the plane of the paper of FIG. 5. The line of light 523 strikes an object surface, which at a first distance from the projector is object surface 510A and at a second distance from the projector is object surface 510B. The line of light 523 intersects the object surface 510A (in the plane of the paper) at a point 526, and it intersects the object surface 510B (in the plane of the paper) at a point 527. For the case of the intersection point 526, a ray of light travels from the point 526 through the camera lens perspective center 537 to intersect the photosensitive array 2641 at an image point 2646. For the case of the intersection point 527, a ray of light travels from the point 527 through the camera lens perspective center 537 to intersect the photosensitive array 2641 at an image point 647. By noting the position of the intersection point relative to the position of the camera lens optical axis 536, the distance from the camera (and projector) to the object surface can be determined using the principles of triangulation, which typically rely on the “baseline” distance between the perspective centers of the projector 520 and the camera 540. The distance from the projector to other points projected by the line of light 523 onto the object, that is points on the line of light that do not lie in the plane of the paper of FIG. 5, may likewise be found using the principles of triangulation.

In the embodiment of FIGS. 6, 7, the end assembly 600 is coupled to an LLP 605 by a first accessory interface 650 and a second accessory interface 655. In an embodiment, the latch arm 660 is rotated to allow the coupling assembly 650, 655 to lock the LLP 605 in place, thereby connecting the LLP 605 to the end assembly 600 both electrically and mechanically. The LLP 605 includes a projector 610 and a camera 620.

In an embodiment, an accessory noncontact 3D measuring device 800 may be attached to the AACMM 10 as illustrated in FIGS. 8, 9 or detached from the AACMM as illustrated in FIG. 10. In FIG. 8, the noncontact 3D measuring device 800 is attached to the AACMM 10, which further includes a probe tip 92 for contact 3D measurement. In an embodiment, the device 800 is attached to the first accessory interface 650. FIG. 9 shows elements in the device 800, including device body 810, first camera 820A, second camera 820B, and projector assembly 850. In an embodiment, the projector assembly includes two illuminators that project planes of laser light.

FIG. 10 shows the noncontact 3D measuring device, such as a line scanner 1000, detached from the AACMM 10. The noncontact 3D measuring device 800 includes the cameras 820A, 820B, and projector assembly 850 described in FIGS. 8, 9. It further includes a handle 1010 and optional light-emitting diodes (LEDs) 822A, 822B.

FIG. 11 shows the noncontact 3D measuring device 1000 in a mode of operation in which a plane of laser light is emitted from each of light sources 1110A, 1110B. In an embodiment, each of the light sources 1110A, 1110B emits light at a different wavelength. In an embodiment, the camera 820A has an optical coating that passes the wavelength of the light 1110A and blocks the wavelength of the light 1110B. In contrast, the camera 820B has an optical coating that passes the wavelength of the light 1110B and blocks the wavelength of the light 1110A. In an embodiment, both cameras 820A, 820B pass the wavelengths emitted by the LEDs 822A, 822B so that markers illuminated by the LEDs 822A, 822B are visible to both cameras.

FIG. 12 shows several possible accessories that may be used with the 3D measuring device 1000. In an embodiment, the 3D measuring device attaches to a wearable unit 1200 that includes a computing unit 1205 and a battery 1210. In an embodiment, the battery 1210 is rechargeable and removable. In an embodiment, the wearable unit receives a signal over a USB or Ethernet cable 1215. Ethernet is a family of computer networking technologies first standardized in 1985 as IEEE 802.3. Ethernet that supports 1 gigabit per second is often referred to as Gigabit Ethernet. Higher speed Ethernet versions with multi-gigabit bandwidth such as 2.5G, 5G, and 10G are becoming increasingly common. In embodiments, the cable 1215 carries one of Gigabit Ethernet, 2.5G, 5G, and 10G. First released in 1996, the USB standard is maintained by the USB Implementers Forum. There are several versions of USB from the initial USB 1.0 that operates at 1.2 Mbps to USB4 that operates at 40 Gbps, with intermediate versions having intermediate data rates. Data may be sent from the wearable unit 1200 to an external computer 1220, which might be a desktop computer or a computer network. Connection from the wearable unit 1200 may be made through cable 1230, through wireless connection 1235, or through a removable memory storage device. Connection may alternatively be made between the 3D measuring device 1000 and the external computer 1220.

Captured data may be displayed using a mobile display unit 1240. In an embodiment, the mobile display 1240 is magnetically attached to the rear side of the 3D measuring device 1000. The mobile phone may receive power from the 3D measuring device 1000, which in turn may receive power from the battery 1210 or external computer 1220. The mobile display 1240 may communicate with the wearable unit 1200 through wireless connection 1245 or through a cable from the wearable device. Alternatively, captured data may be displayed using a monitor 1222 provided to operate in conjunction with the external computer 1220.

FIG. 13A shows a handheld 3D measuring device 1300 (e.g., a photogrammetric camera or line scanner) in which a shaft 1305 provides a handle for an operator 1302. The 3D measuring system 1300 illustrated in FIG. 13A may be operated in a target tracking mode or a geometry tracking mode, according to a selection made by the operator. FIG. 13A illustrates features applicable to both modes.

FIGS. 13A, 13B illustrate the target tracking mode. In this mode, light source 1310A emits a plane of light at the first wavelength. This light is captured by the camera 1320A as the line 1330A. At the same time, light source 1310B emits a plane of light at a second wavelength. This light is captured by the camera 1320B as the line 1330B. In an embodiment, the first wavelength is different than the second wavelength. At the same time, LEDs 1322A, 1322B emit light at a different third wavelength to illuminate reflective markers 1330C placed on or near the object under test. The first camera 1320A includes optical elements coated to pass the first and third wavelengths, while the second camera 1320B includes optical elements coated to pass the second and third wavelengths. Hence each of the cameras 1320A, 1320B sees one of the two projected lines of laser light as well as the illuminated reflective markers 1320C. The lines of light imaged by the cameras 1320A, 1320B are processed to determine the 3D coordinates of illuminated points on the object within the frame of reference of the 3D measuring device 1300. The reflective 3D markers 1330C imaged by the cameras 1320A, 1320B are processed to determine the 3D coordinates of the markers 1330C in successive frames. This enables the 3D coordinates determined for the lines 1330A, 1330B to be tracked (registered) over successive frames.

In the geometry tracking mode illustrated in FIG. 13C, light source 1312A emits multiple parallel planes of light 1340 at a fourth wavelength. The fourth wavelength is different than the first wavelength, second wavelength, and third wavelength. The first camera 1320A and the second camera 1320B both include elements coated to pass the fourth wavelength, and hence both cameras 1320A, 1320B see the projected lines 1340A. Because the optical axis of the camera 1320A is more closely aligned to the optical axis of the projector 1312A than to the optical axis of the projector 1312B, the projected lines of light from the projector 1312A will tend to sweep more slowly across the image sensor as the distance to the object changes than will the projected lines of light from the projector 1312B. The difference in these lines of light as seen by the cameras 1320A, 1320B enables the identity of each line to be uniquely determined. The process of identifying which projected lines correspond to which imaged lines is referred to a “disambiguation” of the lines. In an embodiment, a method used for doing this disambiguation is described in Willomitzer et al., “Single-shot three-dimensional sensing with improved data density,” in Applied Optics, Jan. 20, 2015, pp 408-417. Further improvement in the geometry tracking mode is possible by further projecting multiple planes of light 1340B with the projector 1312B. In an embodiment, the patterns 1340A, 1340B are alternately projected.

As illustrated in FIGS. 13C, 13D, the projected multiple planes of light appear as lines of light 1340A, 1340B when striking a planar surface. Deviations in the imaged lines of light 1340A, 1340B from perfect straightness indicates that the surface being measured is not perfectly planar. Deviations resulting from edges, dips, or bulges can be detected and correlated from shot to shot to determine the amount and direction of movement in each frame. An advantage of the geometry tracking mode compared to target tracking mode is faster measurements since adhesive markers do not need to be applied or removed.

In the embodiment illustrated in FIG. 13E, the photogrammetric camera 1300 is powered by a battery 1210 within a wearable unit 1200. In an embodiment, the power connector 1216 is conveniently disconnected from a handheld scanner such as the scanner 1000, 1300 and plugged into the scanner handle to provide power to the photogrammetric camera. In an embodiment, computing unit 1205 is used to process images obtained by the photogrammetric camera 1300 of target markers affixed on or near the object under test. Computing unit 1205 may also cooperate with an external or networked computer 1220 to process target images. In an embodiment, the mobile display 1240 is used to provide instructions or information on preferred positions and orientations of the photogrammetric camera 1300 in capturing images. In addition, in an embodiment, the mobile display 1240 displays captured data using the mobile display unit 1240. In an embodiment, the mobile display 1240 is magnetically attached to the rear side of the 3D measuring device 1300. The mobile phone may receive power from the 3D measuring device 1300, which in turn may receive power from the battery 1210 or external computer 1220. The mobile display 1240 may communicate with the wearable unit 1200 through wireless connection 1245 or through a cable from the wearable device. Alternatively, captured data may be displayed using a monitor 1222 provided to operate in conjunction with the external computer 1220.

A photogrammetric camera 1350 shown in FIG. 13F may be used in combination with a handheld line scanner such as the scanner 1300. The photogrammetric camera 1350 includes a camera assembly 1360, which include a camera lens, image sensor, and electronics. Surrounding the camera assembly 1360 are a collection of light sources 1370 such as light emitting diodes (LEDs). In an embodiment, the photogrammetric camera further includes a handle 1380 having control buttons 1382, 1384. In an embodiment, the photogrammetric camera is used with scale bars or other scaled objects to provide scale in the captured images. In an embodiment, the light sources 1370 illuminate the object, which may include target reflectors or markers 1330C like those shown in FIG. 13B. Markers 1330C may also be placed on the scale bars. In an embodiment, markers 1330C are placed over a relatively large area on the object. The photogrammetry camera 1350 captures images of the object and scale bars from a variety of positions and perspectives. Software is then used to perform a least-squares fit (or other optimization procedure) to determine the 3D coordinates of the markers in space over the relatively large area of the object. This enables the handheld line scanner 1300, which may measure over a relatively small area at a time, to be accurately registered over a much larger area. If the photogrammetric camera 1350 is used with the scanner 1300 in the geometry tracking mode illustrated in FIGS. 13C, 13D, the photogrammetric camera may be used to measure natural features such as edges or corners to provide registration assistance for a handheld line scanner such as the scanner 1300.

In the embodiment illustrated in FIG. 13F, the photogrammetric camera 1350 is powered by a battery, which may for example be inserted into the handle 1380. In an alternative embodiment illustrated in FIG. 13G, the photogrammetric camera 1350 is powered by a battery 1210 within the wearable unit 1200. In an embodiment, the power connector 1216 is conveniently disconnected from a handheld scanner such as the scanner 1000, 1300 and plugged into the handle 1380 to provide power to the photogrammetric camera. In an embodiment, computing unit 1205 is used to process images obtained by the photogrammetric camera 1350 of target markers affixed on or near the object under test. Computing unit 1205 may also cooperate with an external or networked computer 1220 to process target images. In an embodiment, the mobile display 1240 is used to provide instructions or information on preferred positions and orientations of the photogrammetric camera 1350 in capturing images.

FIG. 14 is a block diagram illustrating exemplary electronics 1400 within a handheld line scanner such as the handheld line scanner 1000 or 1300. Processing for images captured by each of the two image sensors 1410A, 1410B within the line scanner is carried out by a corresponding field programmable gate arrays (FPGAs) 1415A, 1415B and double date rate 4 synchronous dynamic random-access memory (DDR4 SDRAM or simply DDR4) 1418A, 1418B. Printed circuit boards (PCBs) 1420A, 1420B provide direct current (DC) electrical power to components in the electronics 1400. For example, voltages may be provided at 0.9, 1.2, 1.8, 2.5, 3.0, 3.3 and 5 volts. Laser drivers 1430 provide current to lasers 1432 or other light sources that emit lines or other patterns of light. LED drivers 1434 provide current to LEDs ring PCBs 1436. Interface PCB 1440 provides an electrical interface to components outside of electronics 1400. The PCBs 1420A, 1420B also provide electrical power to the button PCB 1442, status LEDs 1444, inertial measurement units (IMUs) 1550, buffers/translators 1452, temperature sensors 1454, and fans 1456. An environmental recorder 1460 records environmental events and is supplied electrical power by battery 1462 to record such events even when power from AC power mains is not available to electronics 1400. For example, the environmental recorder may record high-g shocks measured by the IMUs 1450 during shipping.

FIG. 15 shows electronics 1500 within the exemplary wearable unit 1200 (FIG. 12). In an embodiment, a handheld 3D measurement device such as 1000 or 1300 sends data over a USB-C cable 1505, which can transfer data at up to 10 Gbps to the wearable unit 1200. Data arrives at a first industrial USB connector 1515A within a power distribution PCB 1510. The data is transferred to the USB hub 1520, which in an embodiment is a USB 3.2 Gen 2 hub capable of transferring data at 10 Gbps. Electrical power is delivered to the USB hub 1520 from a battery charger 1542 (via DC/DC converter 1523 for example) that may receive electrical power from either a 19-volt DC line 1540 or from either of two batteries 1545. In an embodiment, the batteries 1545 are removable and rechargeable. The battery charger 1542 sends some DC power to the USB hub 1520, which distributes DC power upstream to the handheld unit (such as 1000 or 1300) according to the instructions of the power controller 1522. The battery charger 1542 also sends some DC power downstream through the DC power output connector 1552 through the cable 1527 to the DC power input connector 1554, which distributes power as required in the components of a System on a Chip (SoC) 1530. Data is passed from the USB hub 1520 to a second industrial USB connector 1515B and through a second USB-C cable 1525 to a USB SuperSpeed+ port 1526 affixed to the SoC 1530. In an embodiment, the SoC 1530 is an Intel Next Unit of Computing (NUC) device. In an embodiment, the SoC 1530 is interfaced to Wi-Fi 1532, Ethernet 1534, and a USB SuperSpeed flash drive 1537. In an embodiment, Wi-Fi 1532 sends wireless signals 1533 to a mobile phone display 1240. Wi-Fi is a trademark of the non-profit Wi-Fi Alliance. Wi-Fi devices, which are compliant with the IEEE 802.11 standard, are used for local wireless network connectivity applications such as to the mobile display 1240 and external computer 1220. In an embodiment, Ethernet 1534 is a Gigabit Ethernet (GbE) that sends signals at 1 Gbit per second over wires (cables) 1535 to an external computer 1220. Ethernet, which is compliant with the IEEE 802.3, is used for wired network connectivity applications. In an embodiment, scan data is saved on a USB SuperSpeed flash drive 1537 via USB port 1536. The Universal Serial Bus (USB) is an industry standard maintained by the USB Implementer's Forum. USB is designed to provide power as well as data communications. USB-C SuperSpeed+ provides data transfer at 10 Gbps. The battery charger 1547 not only delivers DC power from the batteries to the battery charger when required, it also charges the batteries 1545 when power is being supplied by the DC power line 1540.

To improve accuracy of determined 3D coordinates of points measured on an object by a 3D measuring device such as 1000 or 1300, it is desirable to increase the dynamic range of the imaged lines of laser light as much as possible. When dynamic range is large, the 3D measuring device can capture bright reflections without saturation and faint reflections without excessive noise. One method of increasing dynamic range was described in commonly owned U.S. patent application Ser. No. 17/073,923 (hereafter Faro ‘923) filed on Oct. 19, 2020, the contents of which are incorporated by reference herein. This method uses a photosensitive array having a selectable conversion gain (CG), where CG is defined as the voltage produced per electron (e) in a pixel electron well. For example, a pixel having a CG=130 μV/e produces a 1.3-volt signal in response to 10,000 electrons in its electron well. A CG is said to be selectable when any of two or more CGs can be selected. According to one method described in Faro '923, high and low CGs are alternately selected, and the signal obtained for the preferred CG is chosen.

For the electronics illustrated in FIG. 14, a potential disadvantage of the selectable gain method of Faro '923 is that more computations are required of electronics such as the FPGAs 1415A, 1415B. The added computations result in increased power consumption, increased system weight, and added expense to obtain the desired high dynamic range. In an embodiment described in Faro '923, the gain settings are alternated between high gain and low gain, the pixel values are alternately read out, and one of the two read-out values is selected for each pixel. Using this approach, high dynamic range is achieved without increased power consumption, weight gain, or expense.

A method that provides high dynamic range without increasing power consumption, system weight, and expense is illustrated in the method 1600 of FIG. 16. An element 1610 includes, with a 3D measuring device having an image sensor, projecting a pattern of light onto an object. An element 1612 includes, with an image sensor, capturing an image of the projected pattern of light, the captured image having pixel values each based at least in part on a selection among two or more-pixel conversion gains. An element 1614 includes reading out the selected pixel values from the image sensor. An element 1616 includes, with a processor, determining 3D coordinates of points on the object based at least in part on the projected pattern of light and the read-out pixel values.

FIGS. 17A, 17B, 17C illustrate an embodiment of the method 1600 illustrated in FIG. 16. In each of FIGS. 17A, 17B, 17C, the horizontal axis 1702 of each graph represents input data, which is to say the electrical signal (for example, in microvolts) generated in response to electrons in the pixel well. As an example of low and high CG modes, the high CG might be CG_high=130 μV/e while the low CG might be CG_low=30 μV/e. Corresponding numbers of electrons in a full pixel well might then be 10,000 electrons for the high CG case and 40,000 electrons for the low CG case. Corresponding noise levels might be 2 electrons for the high CG case and 9 electrons for the low CG case. In some embodiments, the combining of low CG and high CG within the image sensor is accomplished through the use of a dual-ADC (analog-to-digital converter).

FIG. 17A shows a pixel response curve for the high CG case. For this case, the horizontal axis 1702 may be considered to equivalently represent either the number of photons striking the well or the number of electrons stored in the well. The pixel output data represented by the vertical axis 1704 may be given in voltage. For the case in which light is faint so that relatively few photons reach the pixel well, pixels remain below the saturation limit 1708 while having the advantage of a relatively low readout noise (2 electrons in this example). For the case in which the light level is above the saturation limit 1708, the output response saturates, which is to say that the output voltage of the well levels off to a saturation output level 1710.

FIG. 17B shows a pixel response curve for the low CG case. For this case, the horizontal axis 1712 represents the number of photons striking the well or the number of electrons stored in the well. The pixel output data represented by the vertical axis 1714 may be given, for example, in voltage. For the case in which light is strong so that relatively many photons reach the pixel well, saturation is avoided. Even though the readout noise is relatively larger in this case compared to the high CG case, the signal-to-noise ratio is still relatively good.

FIG. 17C illustrates an embodiment for combining the results of the high CG and low CG data to obtain a high dynamic range (HDR). For this case, the horizontal axis 1702 represents input data and the vertical axis 1724 represents the output data. FIG. 17C illustrates a method for combining the high gain response curve 1706 with the low gain response curve 1716 to obtain a composite response curve that includes an extended region 1726 that results in an HDR response. For input data having a level above the saturation limit 1708, the captured input data 1710 to the right of the saturation limit 1708 is increased by the ratio of high CG to low CG. This causes the input data obtained for the curve 1716 below the saturation limit 1708 to be increased in a movement 1730 by the amount 1740, which when converted to bits is referred to as the bit extension. Since the signal-to-noise ratio is approximately the same for the high CG and low CG, the dynamic range is improved approximately by the bit extension, resulting in HDR. As shown in FIG. 17C, the bit extension 1740 seamlessly extends the range of output values in the extended region 1726 to obtain the HDR.

In another embodiment illustrated in FIG. 18A, image sensors such as the sensors 1410A, 1410B use a method of gradation compression to obtain HDR, enabling a scanning 3D measuring device such as 1000 or 1300 to measure both relatively very dim and very bright reflections. In an embodiment, the image sensors 1410A, 1410B are set to have a plurality of compression break points such as the points/levels 1812, 1822. As in the discussion of FIGS. 17A, 17B, 17C, the horizontal axis 1802 in FIG. 18 represents input data, which is to say the electrical signal (for example, in microvolts) generated in response to electrons in the pixel well. The pixel output data represented by the vertical axis 1804 may also be given in voltage. In an embodiment, for input data between 0 and the level 1812 and an output data between 0 and level 1816, gradation compression is not applied to the input data, resulting in the response curve 1814. For input data in the region between 1812 and 1822, the gain is reduced or compressed, resulting in a smaller slope in the response curve 1824. For input data in the region between 1822 and 1832 (having an output data corresponding to level 1826), the gain is further reduced or compressed, resulting in a still smaller slope in the response curve 1834. The maximum level of the resulting output data is given by the line/level 1840. For example, in a representative image sensor, the level 1840 might correspond to 12 bits (or 4095). Without compression, the signals may be considered small signals covering the range 1818, medium signals that further cover the range 1828, or large signals that further cover the range 1838. In effect, the maximum signal without compression 1836 is compressed to the level 1832. Hence, as illustrated in FIG. 18, the method of gradation compression increases dynamic range.

FIG. 18B describes elements in a method 1850 for using gradation compression to increase dynamic range. An element 1860 includes, with a 3D measuring device having an image sensor, projecting a pattern of light onto an object. An element 1862 includes, with the image sensor capturing an image of the projected pattern of light, the captured image having pixel values based at least in part on a change in pixel response at a plurality of compression break points. An element 1864 includes reading out the selected pixel values from the image sensor. An element 1866 includes, with a processor, determining 3D coordinates of points on the object based at least in part on the projected pattern of light and the read-out of pixel values.

As illustrated in FIG. 4, in a typical case, an emitted laser line 400 is usually projected perpendicular to a line connecting the projector 210 to the camera