Physical-object scanning and measuring systems based on photogrammetry, apparatuses incorporating such systems, and related methods

Abstract

3D scanners that construct 3D virtual scaled models from images of an actual 3D object acquired from multiple perspectives at differing image-capture points throughout a virtual image-capture mesh. In some embodiments, the 3D scanners make measurements of the actual 3D object virtually by applying virtual calipers to the 3D virtual scaled model. In some embodiments, a 3D scanner of this disclosure is deployed within a testing chamber, such as a testing chamber for testing battery devices, such as battery packs, battery modules, battery cells, and the like. In such deployments, the 3D scanners can make accurate measurements of the battery devices in a safe manner. Methods of building 3D virtual scaled models, of measuring actual 3D objects virtually, and of testing battery devices, among others, are also disclosed.

Description

PHYSICAL-OBJECT SCANNING AND MEASURING SYSTEMS BASED ON PHOTOGRAMMETRY, APPARATUSES INCORPORATING SUCH SYSTEMS, AND RELATED METHODSFIELD OF THE INVENTION

[0001] The present invention generally relates to the field of photogrammetry. In particular, the present invention is directed to physical-object scanning and measuring systems based on photogrammetry, apparatuses incorporating such systems, and related methods.BACKGROUND

[0002] Many endeavors require obtaining measurements of a physical object, such as creating a copy of the physical object, creating a digital model of the physical object, and testing a secondary battery device, among many others. Regarding testing of battery devices, lithium-metal battery technology continues to advance. While lithium-metal batteries have many benefits, including relatively high gravimetric and volumetric energy densities, lithium-metal secondary batteries present challenges in design and construction. Several of these challenges stem from the anodes of lithium-metal cells being of the plating-stripping type in which lithium metal ions plate onto the anodes during charging and strip away from the anodes during discharging. The plating and stripping not only cause significant dimensional changes in the stacking direction of the cells during cycling, but when the electrolyte is not in the solid state, also can result in the formation of mossy lithium and / or dendrites on the anodes.

[0003] Designing and testing of lithium-metal battery packs involve measuring battery pack dimensions at differing stages of charging - discharging cycling over the cycle lives of the battery packs. However, taking multipoint measurements manually is tedious and time consuming and impossible for difficult-to-reach points. Additionally, it can be risky to technician safety, especially after a battery pack with a new untested chemistry / design has already been cycled many times and could be developing a condition, such as a dendrite-induced internal electrical short and / or electrolyte decomposition, that could lead to an explosion or other catastrophic failure.SUMMARY OF THE DISCLOSURE

[0004] In one implementation, the present disclosure is directed to a method of generating a three-dimensional (3D) virtual scaled model of a 3D object. The method includes receiving a plurality of images of the 3D object taken from a plurality of differing poses at corresponding image-capture points that form an image-capture mesh that surrounds at least a portion of the 3D object; receiving pose data for each of the images; digitally processing the images to identify physical features of the 3D object captured in ones of the images as captured physical features; digitally building a 3D point-cloud representation of the captured physical features using the acquired pose data and the captured physical features; and digitally scaling the 3D point-cloud representation as a function of physical dimensions of the image-capture mesh in creating the 3D virtual scaled model of the 3D object.

[0005] In another implementation, the present disclosure is directed to a method of making a dimensional measurement of a 3D object. The method includes performing the method described above to create the 3D virtual scaled model of the 3D object; selecting a location on the 3D scaled digital model at which to measure the dimension; and applying a caliper algorithm to the selected location on the 3D virtual scaled model, wherein the caliper algorithm is configured to calculate the dimensional measurement at the selected locations.

[0006] In yet another implementation, the present disclosure is directed to a machine-readable hardware storage medium containing machine-executable instructions for performing either of the methods described above.

[0007] In still another implementation, the present disclosure is directed to a method of testing a battery device having a stacking direction along which a plurality of anodes and cathodes are stacked. The method includes automatically cycling the battery device through a plurality of chargedischarge cycles; automatically collecting cycling data on the battery device during the automatic cycling; intermittently performing a measurement method, such as the measurement method described above, a plurality of times so as to obtain at least one dimensional measurement in a direction parallel to the stacking direction; and automatically collecting each at least one dimensional measurement in association with a corresponding portion of the cycling data.

[0008] In still yet another implementation, the present disclosure is directed to a scanner that generates a three-dimensional (3D) virtual scaled model of a 3D object. The scanner includes an object region that contains the 3D object during use of the scanner to scan the 3D object; at least one imaging device that captures a plurality of images of the 3D object from differing poses at corresponding points that form an image-capture mesh that surrounds at least a portion of the object region; a computing system that includes: hardware memory containing machine-executable instructions and storing the images that the at least one imaging device captures; and at least onedigital processor in operative communication with the hardware memory and configured to execute the machine-executable instruction stored in the hardware memory, wherein the machine-executable instructions include machine-executable instructions for: digitally processing the images to identify physical features of the 3D object captured in ones of the images as captured physical features; digitally building a 3D point-cloud representation of the captured physical features using the acquired pose data and the captured physical features; and digitally scaling the 3D point-cloud representation as a function of physical dimensions of the image-capture mesh so as to create the 3D virtual scaled model of the 3D object.

[0009] In another implementation, the present disclosure is directed to a testing system for testing a battery device having a stacking direction along which a plurality of anodes and cathodes of the battery device are stacked. The testing system includes a testing enclosure for containing and supporting the battery device therein; a scanner as described immediately above, wherein the 3D object comprises the battery device; and a testing controller configured to perform a method as described above.BRIEF DESCRIPTION OF THE DRAWINGS

[0010] For the purpose of illustration, the accompanying drawings show aspects of one or more embodiments of the disclosure. However, it should be understood that the scope of this disclosure is / are not limited to the precise arrangements and instrumentalities shown in the drawings, wherein:

[0011] FIG. 1 A is a partial isometric view / partial block diagram of an example three- dimensional (3D) scanner made in accordance with the present disclosure;

[0012] FIG. IB is an enlarged bottom-up view of the top wall of the example testing chamber with which the 3D scanner of FIG. 1 A is deployed, showing components of an example circumferential-drive mechanism for moving the camera to differing image-capture points;

[0013] FIG. 1C is an enlarged partial elevational view of wires / cables extending through the wire ducts of the testing chamber of FIGS. 1A and IB;

[0014] FIG. ID is an enlarged isometric view of the controller assembly of the example 3D scanner of FIG. 1A, showing the single board computer (SBC) used to control some operations of the 3D scanner;

[0015] FIG. IE is an enlarged isometric view of the vertical imaging-device drive system of the example 3D scanner of FIG. 1A;

[0016] FIG. IF is a perspective view of the virtual image-capture mesh that surrounds the 3D object of FIG. 1A;

[0017] FIG. 2A is a graph of top-point accuracy versus number of scans for an example 3D scanner;

[0018] FIG. 2B is a graph of middle-point accuracy versus number of scans for the example 3D scanner of FIG. 2A;

[0019] FIG. 2C is a graph of bottom-point accuracy versus number of scans for the example 3D scanner of FIGS. 2A and 2B;

[0020] FIG. 3 is a representation of a point cloud generated by a 3D scanner of the present disclosure during scanning operations;

[0021] FIG. 4 is a high-level block diagram of an example battery-device testing system made in accordance with the present disclosure; and

[0022] FIG. 5 is a flow diagram illustrating an example method of generating a 3D virtual scaled model of a 3D object in accordance with aspects of the present disclosure.DETAILED DESCRIPTION

[0023] The entire contents of the appended example claims is incorporated into this Detailed Description section as if originally presented herein.

[0024] In some aspects, the present disclosure is directed to methods of generating three- dimensional (3D) virtual scaled models of real-world physical 3D objects (hereinafter, simply “3D objects”). In some embodiments, such a method includes acquiring a plurality of images of a 3D object taken from a plurality of differing poses and processing the acquired images to build a 3D point cloud representing the 3D object and then digitally scaling the 3D point cloud to create a 3D virtual scaled model of the 3D object. In some embodiments, the poses of the image-capturing device(s) at which images are acquired may be considered to define image-capture points of a virtual 3D image-capture mesh (hereinafter “image-capture mesh”) that surrounds or partially surrounds the 3D object being scanned. The image-capture mesh may be any suitable shape, such as, but not limited to, cylindrical, hemispherical, and ellipsoidal, among others. In some embodiments, the image-capture mesh may be, for example, predefined (e.g., set during manufacture), user-defined (e.g., using a mesh-setup procedure or a selection menu), or automatically defined (e.g., according to the object being scanned), or any desired one of these in a user-selectable manner.

[0025] Each 3D object may be any real- world physical object of any size and shape that can be captured in a plurality of images acquired at differing poses around the 3D object. For the sake of simplicity, the term “3D object” includes not only a single unitary object, but also a cluster or other grouping of multiple individual unitary 3D objects. Those skilled in the art will readily appreciate that the type and size of a 3D object suitable for scanning in accordance with the present disclosure is limited only by limitations on the scanning imposed by the design of a particular 3D scanner, such as size of the image-capture mesh and the modality(ies) used to capture the images. The size of the image-capture mesh may be small (e.g., enclose a volume of about 1 cubic centimeter (cc) or less), average (e.g., enclose a volume of about 1000 cc to about 1.5 m3), or large (e.g., enclose a volume of about 10 m3and above), and anything in between.

[0026] Imaging modalities that can be used to capture images during 3D scanning in accordance with the present disclosure include, but are not limited to, visible-light cameras, thermal cameras, light microscopes, scanning microscopes, scanning line cameras, scanning radars, and scanning sonars, among others, and each corresponding captured image may be a 2D image or a 3D image, as the particular modality allows. In some embodiments, a single image-capture device may be used at every image-capture point of the image-capture mesh, for example, by moving the image-capture device relative to the 3D object or moving the 3D object relative to the image-capture device, or a combination of these two. In some embodiments, a plurality of image-capture devices may be deployed to cover a subset of the image-capture points or all of the image capture points. As an example of the former in the context of a vertically cylindrical image-capture mesh, a set of imagecapture devices may be arranged in fixed relation to one another along a vertical line having a length equal to the height of the cylinder. To capture all of the needed images, the set of image-capture devices can be moved to differing image-capture locations circumferentially around the 3D object. As another example in the context of a vertically cylindrical image-capture mesh, a set of imagecapture devices may be arranged circumferentially around the 3D object, and all of the needed images captured by moving the set of imaging-capture devices to differing vertical locations of the image-capture mesh. Many other arrangements of multiple image-capturing devices are possible for both a vertical cylindrical image-capture mesh and image-capture meshes of other shapes.

[0027] Imaging-device pose data may be in any suitable form and may be determined in any suitable manner. For example, pose data may be based on stepper-motor data, linear-position-sensor data, angular-position-sensor data, global-positioning-system (GPS) data, visual-reference data acquired via the imaging device(s) that capture images of the 3D object, and visual-reference dataacquired via one or more imaging devices other than the imaging device(s) that capture images of the 3D object, among others, and any combination thereof.

[0028] In some embodiments, a point-cloud representation of the 3D object is built by stitching- together information from the acquired images using physical features of the 3D object that are captured in the images (i.e., “captured physical features”) as registration markers for stitching the acquired images together with one another. Each captured physical feature that the method can use as a registration marker for building a point-cloud representation can be any physical feature of the 3D object or environment surrounding the 3D object that digital-image processing and handling can capture and discern and use as a registration marker. A small set of examples of suitable physical features includes, but is not limited to, edges, openings, corners, protrusions, depressions, surface markings, grooves, channels, differences in color, visually distinct regions, tactilely distinct regions, and fixed environmental objects adjacent to the 3D object during scanning, among many others. Fundamentally, the only requirement for a physical feature is that it can be discerned, using imageprocessing software, in two or more of the captured images such that image-stitching software can stitch the captured images together using the physical feature as represented in each of the captured images as matching registration markers. Image-processing software for identifying registration markers and image-stitching software for stitching multiple images together with one another are known in the art and can be readily adapted to a method of the present disclosure. An example of software that can be used to construct a 3D point cloud representation of the 3D object is Meshroom 3D reconstruction software available through the AliceVision Association.

[0029] The point-cloud representation built using the captured images may be considered a 3D virtual model of the 3D object. In some embodiments, it may be necessary or desirable to make the 3D virtual model a 3D virtual scaled model so that the virtual dimensions of the 3D virtual scaled model are in-scale relative to the scanned 3D object, such as in a 1 : 1 scale, a 2: 1 scale, or a 0.5: 1 scale, among many others. In some embodiments, the scaling of the digital 3D model may be accomplished using physical dimensional data for the image-capture mesh used to capture the images of the 3D object and, in some embodiments, data (e.g., magnification data, focal-length data, etc.) regarding the digital size of the 3D object in the captured images on the one hand and the physical size of the 3D object on the other hand. In an example involving a vertical cylindrical image-capture mesh, the size of image-capture mesh is known based on user input of a maximum height of the cylinder and a radius or diameter of the cylinder. The point locations, i.e., the imagingdevice poses, and “image-capture points”, on the image-capture mesh are known (e.g., by height andangle about the circumference of the cylinder and from the number of radial and vertical locational stops to acquire the images), and therefore the spacing between the image-capture points is also known.

[0030] After constructing a point-cloud representation of the 3D object and the image-capture points of the image-capture mesh, for example, using the Meshroom software mentioned above, the spacing in the point cloud between the image-capture points are compared with the real-world spacing of the image-capture points where the images were actually acquired (which are constant). The point cloud is then scaled based on that comparison.

[0031] For orientation and in an example illustrated by FIG. IF, a plane (not illustrated) is fitted to a lower layer of the image-capture points 128CP (such as the plane containing the bottom-most circle of the image-capture points in FIG. IF), for example, using a random sample consensus (RANSAC) fitting algorithm. Then, the point cloud representation of the scanned 3D object is rotated based on the Euclidean rotation vector of the plane being obtained by the RANSAC fitting algorithm.

[0032] In some embodiments, it is desirable to take measurements of the 3D object digitally using the 3D virtual scaled model. In one example, such measurements may be made using a virtual caliper algorithm that generally mimics a physical caliper that a user could use to measure the actual physical 3D object. An example of a virtual caliper algorithm is described below in the DETAILED EXAMPLES section.

[0033] In some aspects, the present disclosure is directed to a 3D scanner that includes the appropriate hardware and software to perform any of the methods discussed above and / or addressed by the appended example claims. A detailed example of such a 3D scanner is described below in the DETAILED EXAMPLES section, and other example features of a scanner of the present disclosure are addressed in the appended claims.

[0034] In some aspects, the present disclosure is directed to a testing apparatus, such as a battery-device testing apparatus, that includes a 3D scanner of the present disclosure. In some embodiments, the 3D scanner of such a testing apparatus may be used to take virtual measurements of a 3D object during testing. This can be especially useful in applications in which taking physical measurements manually can be difficult and / or risk the safety of the person taking the measurements. An example of such risk is described above in the Background section.

[0035] Detailed examples of such battery-device testing apparatuses and systems are described below in the DETAILED EXAMPLES section, and other example features of battery-device testing apparatuses of the present disclosure are addressed in the appended claims. A battery device may be, for example, a battery pack, a battery module, a battery, a group of battery cells, or a battery cell, among other things. The battery device may be based on any suitable electrochemistry, such as, but not limited to, lithium (Li)-metal-battery electrochemistry and Li-ion-battery electrochemistry, among others.

[0036] Aspects related to battery-device testing apparatuses of the present disclosure include methods of operating such battery-device testing apparatuses. Examples of such methods are provided in the DETAILED EXAMPLES section below, as well as in the appended claims.

[0037] DETAILED EXAMPLES

[0038] FIGS. 1 A through IF illustrate an example 3D scanner 100 of the present disclosure and various components of the scanner in the context of a testing chamber 102 for testing battery packs. In this connection and as discussed above, it is noted that a 3D scanner of the present disclosure need not be associated with a testing chamber. For example, such a 3D scanner can be completely independent of other equipment or integrated with or supplemental to some other type of equipment than a battery-pack tester. In this embodiment, the 3D scanner 100 uses photogrammetry techniques to create a 3D virtual scaled model 104 of a 3D object 108, here a battery pack, placed in the middle of a testing chamber 102. Then, the 3D scanner 100 makes virtual measurements of the 3D object 108 by mimicking the way physical calipers are used to take dimensions of an actual 3D object and applying the “virtual caliper” to the 3D virtual scaled model 104. To create the virtual caliper, two parallel planes 104PP(l) and 104PP(2) are created in the 3D virtual scaled model 104 at the surfaces of the two opposite faces 104F(l ) and 104F(2) being used to calculate thickness of the 3D object 108. Then a distance formula ( d = T / (X2— x-^2+ (y2— y^)2+ (z2— z^2) is used to calculate the distance between corresponding ones of the corners and centers and the planes 104PP(l) and 104PP(2).

[0039] To obtain a successful 3D reconstruction of the 3D object 108 by way of the 3D virtual scaled model 104, the 3D scanner 100 acquires a series of images 116 (FIG. 1 A) around the 3D object. In this example, camera-pose mechanisms 120 move a visible-light camera 124 both radially and vertically to image-capture points 128CP (only a few labeled to avoid clutter) (FIG. IF) on a cylindrical image-capture mesh 128 that laterally surrounds the full body of the 3D object 108. Theacquired images 116 (FIG. 1 A) are provided to a computing system 132 that executes a suitable 3D- model-building algorithm 136 that creates a virtual representation of the 3D object 108 based on distinctive features of the 3D object captured in the images, such as characteristics and patterns, for example, alpha-numeric characters that may be present on the 3D object, colors, damages (e.g., scrap marks, dents, scratches, etc.), or any non-repetitive mechanical shape on and / or surrounding the 3D object being scanned.

[0040] Photogrammetry is based on perspective geometry. When a camera, such as the visible- light camera 124 of FIG. 1A, is calibrated, i.e., the focal length is known, angles between the light rays from various points can be determined. Given at least 3 points in a scene with known 3D coordinates in an arbitrary coordinates system, the position of the perspective center and the orientation of the image with respect to that coordinate system can be determined. Knowing the positions of repetitive instances of physical features across several acquired images (2+), the 3D- model-building algorithm 136 places a point representing each repeating physical feature in the coordinate system, and the 3D-model-building algorithm repeats the process until a point-cloud representation 140 of the 3D object 108 is obtained. In some embodiments, a constant focal length is necessary to establish a proper representation of each characteristic being used for the 3D reconstruction.

[0041] Once the computing system 132 has reconstructed the 3D object 108 into a virtual representation, a scaling-and-orienting algorithm 144 scales and orients the virtual representation using known camera-position data. In an example involving a vertical cylindrical image-capture mesh 128 (FIG. IF), before acquiring the images 116 for the first time a user specifies the maximum height 128H the visible-light camera 124 shall travel vertically (maximum height) when acquiring the images. Based on the maximum height (6 stops in this example) and a constant radial stops value (32 radial stops (only 14 of which are shown in FIG. IF to avoid clutter and to provide a general representation)) the coordinates where the visible-light camera 124 shall stop to capture images are calculated by the 3D scanner 100. In one example for the battery pack (i.e., the 3D object 108) shown in FIG. 1A, the visible-light camera 124 is positioned at 6 differing equally spaced vertical heights and 32 differing equally spaced horizontal angles to create 192 image-capture points 128CP (camera poses) on the image-capture mesh 128 that surrounds the 3D object. In the present instantiation, the 32 circumferential stops and the 6 vertical stops ensure that at least 40% of each image 116 is being shared by its neighbor images to help the 3D-model-building algorithm 136 to localize the features being used for the 3D reconstruction of the 3D object 108.

[0042] Several tests have been performed during the development process. After a certain number, n, of acquired images, such as images 116, there appears to not be an appreciable benefit in terms of accuracy increment but rather a detriment in terms of processing time. Consequently, the image-capture mesh, such as the image-capture mesh 128, can be designed to strike an appropriate balance between accuracy and processing time if desired.

[0043] The known coordinates of the image-capture points 128CP are then compared to the 3D reconstructed positions / coordinates (which in this example are given arbitrary values by the 3D- model-building algorithm) of the visible-light camera 124. The scaling-and-orienting algorithm 144 then scales and rotates the virtual representation based on the values obtained by the noted comparison so as to create the 3D virtual scaled model 104.

[0044] In the present example, the 3D scanner 100 comprises several main components: the visible-light camera 124 (FIGS. 1A and IE), a stepper-motor driver and a single-board computer (SBC) 152 (FIG. ID), two stepper motors 156(1) and 156(2) (FIG. 1A), a custom-made bearing system 160 (FIGS. 1A and IB) and a wire-management system 164 (FIGS. 1A through 1C). In embodiments in which a 3D scanner of the present disclosure, such as the 3D scanner 100 of FIG. 1A, is integrated with a testing apparatus, such as the testing chamber 102 of FIG. 1A or the testing apparatus 408 of FIG. 4, one or more sensor systems, such as the pressure-monitoring system 460 (FIG. 4) and a temperature-monitoring system 456 (FIG. 4), may be part of the testing apparatus in addition to the relevant components of the 3D scanner 420 of the present disclosure as discussed below.

[0045] FIG. 1A shows some internal structure of the 3D scanner 100 within the testing chamber 102 with which the example 3D scanner is deployed. Main cycler wires 168 connecting the battery pack (3D object 108) to a cycler (not shown, but see cycler 412 of FIG. 4) pass through the middle of the upper wall 102UW of the testing chamber 102 to avoid interfering with scanner wires 172 of the 3D scanner 100 itself. This can prevent undesired movement of the battery pack (3D object 108). FIG. 1C shows a cycler-wire duct 164CWD and a scanner-wire duct 164SWD for the cycler wires 168 and scanner wires 172, respectively, in more detail. To avoid twisting of the scanner wires 172, the 3D scanner 100 may alternate the direction of scanning from one scan to the next (so as to alternatingly “twist” and “untwist” the wires) to ensure reliable electrical connections.

[0046] FIG. IB shows components of a circumferential drive mechanism 120C of the camerapose mechanisms 120 that move the visible-light camera 124 to differing angular positions duringthe scanning process either clockwise or counterclockwise or alternating between the two directions. The circumferential drive mechanism 120C is driven by the stepper motor 156(1) having a gear 156(1)G. The movement of the stepper motor 156(1) is transmitted to a main body 120MB (FIG. 1 A) of the circumferential drive mechanism 120C through the gear 156(1)G that enmeshes with a gear 120Gthat is connected via a rotational bearing 120RB that acts to rotationally connect the gearl20Gto the upper wall 102UW of the testing chamber 102 to establish a vertical rotational axis 176 for the circumferential drive mechanism 120C. A toothed outer periphery of the gear 120G is enmeshed with a gear 156(1)G of the stepper motor 156(1), and the rotational bearing 120RB allows the gear to rotate freely relative to a fixed support 120FS secured to the upper wall 102UW of the testing chamber 102. A wire-management guide 120WG is also attached to the gear 120G so as to be rotatable therewith. This ensures that the scanner- wire duct 164SWD (FIG. 1C) of 3D scanner 100 is always facing the SBC 152 (FIG. ID) when the structure rotates, by rotating the wiremanagement system 164 with the gear 120G. A nut 164N in the wire- management system 164 holds the top part 164TP (FIG. 1C) of the wire-management system with its lower part 164LP (FIG. 1C). The cycler- wire duct 164CWD is essentially a hole through the central region of the wiremanagement system 164 that allows the cycler wires 172 between the cycler and the battery pack (3D object 108) to avoid interacting with the scanner wires 168 going through to the SBC 152 to free them from disturbances caused by the motion of the structure. A bridge connection 120BC is on the side opposite to the wire-management guide 120WG (FIG. IB). This allows the physical connection / joint between the circumferential drive mechanism 120 and the rest of the components of the 3D scanner 100 internal to the testing chamber 102.

[0047] FIGS. ID and IE show, respectively, the SBC 152 and a vertical drive mechanism 120V of the camera-pose mechanisms 120 (FIG. 1A). FIG. ID also shows a bridge 120B that connects the vertical drive mechanism 120V (FIG. IE; see also FIG. 1A) to the circumferential drive mechanism 120C shown in FIGS. 1 A and IB. FIG. IE shows the vertical drive mechanism 120V that moves the visible-light camera 124 up and down under the control of the SBC 152. In this example, the vertical drive mechanism 120V includes the stepper motor 156(2) (FIG. ID) that drives a lead screw 120LS that moves the visible-light camera 124 up and down and two vertical rods 120R(l) and 120R(2) that keep the visible-light camera from pivoting when the stepper motor 156(2) rotates the lead screw.

[0048] It is noted that while stepper motors 156(1) and 156(2) are shown for providing both vertical movement and circumferential movement, other actuators / mechanisms can be used, such asservomotors, belt-type mechanisms, and linear actuators other than screw-type actuators, among others.

[0049] During use of the example 3D scanner 100, once a user has placed a battery pack (the 3D object 108) in the middle of the testing chamber 102, the user may hit a “Scan” button 180B on a user interface (UI) 180 of, for example, 3D scanner or a testing controller (not shown) in operative communication with the 3D scanner. In response, control software (not shown) may send a signal to the SBC 152 that controls the camera-pose mechanisms 120 and the visible-light camera 124. In one example, the SBC 152 controls the camera-pose mechanisms 120 to displace the visible-light camera 124 around the 3D object 108 to the 192 image-capture points, here 11.25° per radial step equivalent to 32 pictures / stops per full radial movement (found to deliver the most optimal results in the present instantiation) and up and down the 6 vertical stops and to cause the visible-light camera to acquire an image 116 of the 3D object at each of the 192 image-capture points. In this example, the 192 images of the 3D object 108 are sent to computing system 132, which can be any suitable computing system, such as a laptop computer, desktop computer, cloud-based computing system, computing system of a testing controller, a computing system aboard the SBC 152 or other processing system integrated into the 3D scanner 100 or other equipment, if any, with which the 3D scanner is used, etc. As those skilled in the art will readily appreciate, fundamentally there are no limitations on the computing system 132 other than it be able to perform the necessary functions discussed herein.

[0050] Then, using photogrammetry techniques as discussed above, the 3D-model-building algorithm 136 constructs the 3D point-cloud representation 140 that is used to make a digital representation of the physical 3D object 108 that is then used to make the 3D virtual scaled model 104. The 3D-model-building algorithm 136 may use the time-stamps at which the visible- light camera 124 acquired the images 116 to organize the images and speed up the reconstruction process by specifying the sequence of the images and their proximity to each other, thereby making the identification of unique features within adjacent ones of the images and / or other similarities between the images less computationally intensive. In the present instantiation, the 3D virtual scaled model is within a ±0.2mm accuracy relative to the actual 3D object, as shown in results in FIGS. 2 A through 2C. In this example, the accuracy was determined by doing a test on a battery pack (3D object 108) involving 62 scans and measurements, and the values obtained via the virtual caliper by the 3D scanner were compared to the actual values measured on the actual 3D object using a physical caliper.

[0051] In an example, the physical locations of the camera poses 128CP on the image-capture mesh 128 (FIG. IF) are determined by the number of steps the stepper motors 156(1) and 156(2) (FIGS. 1A, IB, and ID) have taken to move the visible camera 124, both circumferentially and vertically. Reconstructed camera locations (virtual locations) of the point-cloud representation 140 are determined based on the name of the visible-light camera 124. Every image 116 being taken has a different name. In one instantiation, the nomenclature goes as follows: TestName_cycle_cycle#_SOC_Frame_radialStop#_VerticalStop#. The radial stop # and the vertical stop # give the reference location of the camera. As an example, TestName_cycle_cycle#_ SOC_Frame_00_0 refers to the first image being taken, while TestName_cycle_cycle#_SOC_ Frame_31 6 refers to the last image being taken.

[0052] Since the circumferential and vertical stops of the camera poses 128CP (FIG. IF) are constant for every scan, the space between each camera pose and corresponding image 116 is known (the radius in the example image-capture mesh 128 of FIG. IF) is constant, while the maximum vertical travel distance, i.e., height 128H, the visible-light camera 124 has to reach is defined by the user and limited to the design of the 3D scanner 100 - 0mm to 125mm in this example).

[0053] Once the 3D-model-building algorithm 136 has constructed the point-cloud representation, the 3D-model-building algorithm gives the virtual locations of the visible-light camera (corresponding to the camera poses 128CP) random locations but they share the same characteristics from one to another (being arranged in the same order as the pictures). Each image 116 is reconstructed with data, such as a unique ID and name, by matching the names of the images with the virtual image-capture points representing the corresponding camera poses 128CP with the unique ID so as to identify the ID for each reconstructed visible-light camera. Then the ID for each virtual image-capture point within the point-cloud representation is used to retrieve camera location and rotation (pose) within the reconstruction files. FIG. 3 shows an example actual pointcloud representation 300 (corresponding to point-cloud representation 140 of FIG. 1 A) of a battery pack (not shown) made using an instantiation (not shown) of the 3D scanner 100.

[0054] FIG. 4 illustrates an example battery-device testing system 400 for testing a battery device 404, such as a battery cell, battery module, or battery, among other devices. In this example, the battery-device testing system 400 includes a testing apparatus 408 for containing the battery device 404 during testing, a cycler 412 for cycling the battery device, and a computing system 416 for performing and / or overseeing the control of any one or more of the various systems andsubsystems of the battery-device testing system 400. The testing apparatus 408 incorporates some components of a 3D scanner 420 made in accordance with the present disclosure, such as any 3D scanner discussed above generally and, in one instance, the 3D scanner 100 described above in connection with FIGS. 1 A through IF. In this connection, it is noted that existing 3D-scanner solutions available in the market need post-processing work, such as manually scaling the virtual 3D object to real- world dimensions using one or more other physical objects as a scale reference. In contrast, a testing apparatus of the present disclosure, such as the testing apparatus 408 of FIG. 4, avoids such additional work by using the positions / coordinates of the images (not shown) that the 3D scanner, for example, 3D scanner 420, acquires to virtually reconstruct the 3D object, here the battery device 404.

[0055] The testing apparatus 408 includes a testing chamber 408C that contains the battery device 404 as well as certain components of the 3D scanner 420, such as one or more cameras (singly and collectively represented at 420C) and any mechanism(s) (singly and collectively represented at 420M) needed to move the camera(s) to the various camera poses. Additional components, such as local controller(s) (see, e.g., SBC 152 of FIG. ID) and microprocessor(s), among others, may or may not be located within the testing chamber 408C in accordance with a particular design. Components and / or functionality of the 3D scanner 420 not located within the testing chamber 408C, such as main control and processing, may be located elsewhere, such as aboard the computing system 416.

[0056] In this example, the computing system 416 may be composed of any one or more computing architectures, such as a local architecture (e.g., a local server, a laptop computer, a desktop computer, a custom computing device, etc.), a central architecture (e.g., an enterprise server, a remote server, etc.), or a distributed architecture (e.g., a cloud-type, multisite edge computing type, etc.), among others, and any suitable combination thereof. Fundamentally, there are no limitations of the architecture of the computing system 416 other than it be able to provide the necessary / desirable functions. Generally, the computing system 416 includes, among many other elements commonly known in the art, one or more microprocessors (singly and collectively represented at 416P), a plurality of memories (collectively represented at 416M), and one or more input / output devices (singly and collectively represented at 416D). Each microprocessor 416P may be any suitable microprocessor known in the art. The memory 416M may comprise any and all types of hardware storage memory known in the art for computing systems. Examples of such memory include, but are not limited to RAM, ROM, cache, etc., embodied in any suitable form, such as solidstate electronic, optical, or magnetic, among others. As used herein and in the appended claims, the term “computer-readable storage medium” covers any and all suitable hardware, which necessarily excludes transient signals, such as signals on a carrier wave and pulsed signals that are generated by hardware. In contrast, the term “computer-readable medium” includes such transient signals.

[0057] The memory 416M of the computing system 416 contains all of the computer-executable instructions 4161 needed to, among other things, perform and / or control aspects of the 3D-scanning and battery-device testing not performed by any other component(s) of the battery- device testing system 400, as well as stores any and all data needed for the battery-device testing system 400 to perform all of the necessary / desired functions and any and all data collected and / or manipulated during testing. Examples of computer-executable instructions include, but are not limited to, computer-executable instructions for performing algorithms, for providing an operating system, for controlling communications, for controlling one or more off-board devices and / or systems, and for collecting data, among many others. An example algorithm that may be embodied in the computerexecutable instructions is a 3D-model-building algorithm 424; others are discussed above and below.

[0058] A 3D-model-building algorithm of the present disclosure, such as 3D-model-building algorithm 424, can apply one or more automatic filters based on neighbor points in a resulting 3D point-cloud representation 428. For example, an outlier-point filter 424F that removes noise points outside of the point-cloud representation 428 of the battery device 404 (3D object) may be implemented. In some embodiments, the 3D scanner 420 will perform an autonomous scan base on either time or pulses sent by the cycler 412, as chosen by a user.

[0059] In some embodiments, the cycler 412 and / or the computing system 416 may include a UI (here, UI 416UI provided via the computing system 416) that allows a user to make any one or more of a variety of selections for conducting testing and / or controlling the 3D scanner 420. Dimensions of the battery device 404 can be virtually checked via a 3D virtual scaled model 432 resulting from the point-cloud representation 428 and a virtual- caliper algorithm 436 according to user input, based on time, and / or based on signals / feedback received by the cycler 412. For example, based on design or upon the selection of a dimension-collection mode from a menu 444 of such modes on the UI 416UI, a user could specify to scan the battery device 404 each 20 minutes or by specifying the charging and discharging rate. If based on signal option is preferred / chosen, a scanner-control algorithm 448 in the computing system 416 may control the 3D scanner 420 to perform a scan of the battery device 404 every time the cycler 412 applies a step chosen by the user.For example, the user can specify that the 3D scanner 420 perform a scan every rest step after the battery device 404 is fully charged / discharged by choosing the step number within the sequence. In an example, initial thickness calibration may be done once during a test’s setup process. The camera(s) 420C may be calibrated by capturing images of a checkerboard-like image with squares of known dimensions to find distortion values. Then those distortion values may be stored within configuration files 452 in the memory 416M for the 3D scanner 420.

[0060] The battery-device testing system 400 may include any number of additional systems as necessary / desired. For example, the battery-device testing system 400 may include a temperaturemonitoring system 456, a pressure-monitoring system 460, and / or an environmental-control system 464, among others, each of which may be controlled in any suitable manner, such as via the computing system 416. The temperature-testing system 456, if provided, may include one or more sensors (singly and collectively represented at 456S), such as, but not limited to, one or more thermal imaging sensors deployed within the testing chamber 408C and / or one or more temperature sensors applied to or incorporated into the battery device 404, and any suitable system-enabling hardware, such as, for example, analog-to-digital converters, wiring, and signal-conditioning circuitry, among other things, as well as any suitable system-enabling software. The pressuremonitoring system 460, if provided, may include one or more pressure sensors (singly and collectively represented at 460S) located internally and / or externally to the battery device and any suitable system-enabling hardware, such as, for example, analog-to-digital converters, wiring, and signal-conditioning circuitry, among other things, as well as any suitable system-enabling software. The environmental control system 464, if provided, may include one or more heaters 464H, one or more coolers 464C, any suitable system-enabling hardware, such as, for example, analog-to-digital converters, digital-to-analog converters, wiring, and signal-conditioning circuitry, among other things, and any suitable system-enabling software.

[0061] In addition to benefits discussed above, additional benefits of various embodiments disclosed herein include, but are not limited to:• Components of a 3D scanner can fit inside a testing chamber being used to test a battery device. The 3D scanner can be a permanent part of the testing chamber. The 3D scanner can be retrofit into an existing testing chamber or can be built into a new testing chamber.• Measurements made via the 3D virtual scaled model can be performed essentially in real time relative to the collection of the corresponding temperature, pressure, and / or other data that is occurring when a battery device is undergoing a change in physical dimension.• Testing protocols, algorithms, electrical connections, and safety do not need to be modified to integrate a 3D scanner into the testing. Embodiments of the 3D scanner can work with existing testing platforms.• A 3D scanner can assist in collecting data on temperatures, pressures, and dimension profiles across an entire battery device instead of at only single points, expanding understanding of a battery device’s behavior.• Data acquired by the testing apparatus is ready to use for algorithm and safety protocols development.• No human-to-device interaction is required once the testing is running.

[0062] In some examples, a battery-device testing system of the present disclosure, such as the battery-device testing system 400 of FIG. 4, can be used for device study, safety protocol developments, and deep-learning algorithm training to predict battery device behavior along its cycling life, among other things. By collecting data such as current, voltage, full pressure distribution, temperature distribution, and a full 3D profile / thickness / expansion distribution, an algorithm to predict the future status of a battery device can be developed.

[0063] It is noted that pressure, volume, and temperature are corelated to each other. Additionally, C-rate and voltages affect the temperature of a battery device. For example, with implementation of 2D pressure sensors, an IR camera (temperature distribution), and a 3D scanner (expansion / volume) of the present disclosure, trends and characteristics, such as (pressure concentrations in one spot and non-uniform expansion), that indicate that a battery device is soon to reach the end of life can be determined.

[0064] FIG. 5 illustrates an example method 500 of the present disclosure. In this example, the method 500 generates a 3D virtual scaled model of a 3D object, for example, using a 3D scanner of the present disclosure, such as any of the 3D scanners described above generally or the 3D scanners 100 and 420 of FIGS. 1A through IF and 4. At block 505, a plurality of images of the 3D object are received. The images are taken when one or more imaging devices are corresponding image-capture points of an image-capture mesh that at least partially surrounds at least a portion of the 3D object. At block 510, pose data for each of the images is received. Such pose data may include information concerning locations on the image-capture mesh, among other data / information. At block 515, the images are processed to identify physical features of the 3D objects captured in ones of the images as captured physical features. Examples of physical features and howsuch physical features can be used are described above. At block 520, a 3D point-cloud representation comprising the captured physical features using the captured physical features and the acquired posed data. In an example, the 3D point-cloud representation may be built using a process described above, for example, using a 3D-model-building algorithm, such as either 3D-model- building algorithm 136 and 424 of FIGS. 1A and 4, respectively. At block 525, the 3D point-cloud representation is scaled as a function of physical dimensions of the image-capture mesh in creating the 3D virtual scaled model of the 3D object. In an example, the 3D point-cloud representation is scaled to create the 3D virtual scaled model as described above.

[0065] Various modifications and additions can be made without departing from the spirit and scope of this disclosure. Features of each of the various embodiments described above may be combined with features of other described embodiments as appropriate in order to provide a multiplicity of feature combinations in associated new embodiments. Furthermore, while the foregoing describes a number of separate embodiments, what has been described herein is merely illustrative of the application of the principles of the present invention. Additionally, although particular methods herein may be illustrated and / or described as being performed in a specific order, the ordering is highly variable within ordinary skill to achieve aspects of the present disclosure. Accordingly, this description is meant to be taken only by way of example, and not to otherwise limit the scope of this invention.

[0066] Exemplary embodiments have been disclosed above and illustrated in the accompanying drawings. It will be understood by those skilled in the art that various changes, omissions and additions may be made to that which is specifically disclosed herein without departing from the spirit and scope of the present invention.

Claims

What is claimed is:

1. A method of generating a three-dimensional (3D) virtual scaled model of a 3D object, the method comprising: receiving a plurality of images of the 3D object taken from a plurality of differing poses at corresponding image-capture points that form an image-capture mesh that surrounds at least a portion of the 3D object; receiving pose data for each of the images; digitally processing the images to identify physical features of the 3D object captured in ones of the images as captured physical features; digitally building a 3D point-cloud representation of the captured physical features using the acquired pose data and the captured physical features; and digitally scaling the 3D point-cloud representation as a function of physical dimensions of the image-capture mesh in creating the 3D virtual scaled model of the 3D object.

2. The method of claim 1, wherein the images are visible-light images captured by at least one camera.

3. The method of claim 1, wherein digitally processing the images includes using patternrecognition algorithms to identify the captured physical features.

4. The method of claim 1 , wherein digitally building a 3D point-cloud representation includes using a triangulation algorithm and the captured physical features to identify feature points of the point cloud.

5. The method of claim 1, further comprising capturing the images of the 3D object at the corresponding points of the image-capture mesh.

6. The method of claim 5, wherein capturing the images includes moving an imaging device to multiple points of the image-capture mesh.

7. The method of claim 5, wherein capturing the images includes capturing images using multiple imaging devices located at differing points of the image-capture mesh.

8. The method of claim 1, further comprising receiving user instructions that define dimensional parameters of the image-capture mesh.

9. The method of claim 8, wherein the image-capture mesh forms a cylinder, and the 3D object is located within the cylinder.

10. The method of claim 9, wherein the cylinder has a height, and the user instructions include a numerical value for the height.

11. A method of making a dimensional measurement of a 3D object, the method comprising: performing the method of any one of claims 1 through 10 to create the 3D virtual scaled model of the 3D object; selecting a location on the 3D scaled digital model at which to measure the dimension; and applying a caliper algorithm to the selected location on the 3D virtual scaled model, wherein the caliper algorithm is configured to calculate the dimensional measurement at the selected locations.

12. The method of claim 11, wherein the 3D object is a battery device, and the method is performed during testing of the battery device.

13. The method of claim 12, wherein the method is performed multiple times during the testing of the battery device.

14. A machine- readable hardware storage medium containing machine-executable instructions for performing the method of any one of claims 1 through 13.

15. A method of testing a battery device having a stacking direction along which a plurality of anodes and cathodes are stacked, the method comprising: automatically cycling the battery device through a plurality of charge-discharge cycles; automatically collecting cycling data on the battery device during the automatic cycling; intermittently performing a measurement method in accordance with any one of claims 11 through 13 a plurality of times so as to obtain at least one dimensional measurement in a direction parallel to the stacking direction; and automatically collecting each at least one dimensional measurement in association with a corresponding portion of the cycling data.

16. The method of claim 15, wherein the battery device is a battery cell, a battery module, or a battery.

17. The method of claim 15, wherein the cycling data includes one or more types of data selected from a group consisting of electrical-current data, voltage data, state-of-charge data, temperature data, and pressure data.

18. The method of claim 15, wherein intermittently performing the measurement method includes performing the measurement method according to a set schedule relative to the automatic cycling of the battery device.

19. The method of claim 15 or 18, wherein at least one characteristic of the cycling data triggers the performing of the measurement method.

20. The method of claim 15, wherein intermittently performing the measurement method includes performing the measurement method in response to a human input.

21. A scanner that generates a three-dimensional (3D) virtual scaled model of a 3D object, the scanner comprising: an object region that contains the 3D object during use of the scanner to scan the 3D object; at least one imaging device that captures a plurality of images of the 3D object from differing poses at corresponding points that form an image-capture mesh that surrounds at least a portion of the object region; a computing system that includes: hardware memory containing machine-executable instructions and storing the images that the at least one imaging device captures; and at least one digital processor in operative communication with the hardware memory and configured to execute the machine-executable instruction stored in the hardware memory, wherein the machine-executable instructions include machine-executable instructions for: digitally processing the images to identify physical features of the 3D object captured in ones of the images as captured physical features; digitally building a 3D point-cloud representation of the captured physical features using the acquired pose data and the captured physical features; and digitally scaling the 3D point-cloud representation as a function of physical dimensions of the image-capture mesh so as to create the 3D virtual scaled model of the 3D object.

22. The scanner of claim 21, wherein the at least one imaging device comprises a visible-light camera.

23. The scanner of claim 21, wherein the machine-executable instructions for digitally processing the images encode pattern-recognition algorithms to identify the captured physical features.

24. The scanner of claim 21, wherein the machine-executable instructions for digitally building a 3D point-cloud representation encode a triangulation algorithm and the captured physical features to identify feature points of the point cloud.

25. The scanner of claim 21, further comprising a mechanism that moves the at least one imaging device to a plurality of the differing poses at corresponding points of the image-capture mesh, wherein the machine-executable instructions further include machine-executable instructions for controlling the mechanism to capture the plurality of images of the 3D object.

26. The scanner of claim 25, wherein the at least one imaging device is a single imaging device.

27. The scanner of claim 21, wherein the at least one imaging device is multiple imaging devices located at differing points of the image-capture mesh.

28. The scanner of claim 21, wherein the machine-executable instructions encode mesh-defining algorithms that allow a user to define dimensional parameters of the image-capture mesh.

29. The scanner of claim 28, wherein the image-capture mesh forms a cylinder, and the 3D object is located within the cylinder during scanning.

30. The scanner of claim 28 or claim 29, wherein the image-capture mesh has a height, and the mesh-defining algorithms allow the user to define the height of the image-capture mesh.

31. A testing system for testing a battery device having a stacking direction along which a plurality of anodes and cathodes of the battery device are stacked, the testing system comprising: a testing enclosure for containing and supporting the battery device therein; a scanner according to any one of claims 21 through 30, wherein the 3D object of claims 21 through 30 comprises the battery device; and a testing controller configured to perform the method of any one of claims 15 through 20.

32. The testing system of claim 31, further comprising instrumentation and at least one sensor for sensing at least one condition of the battery device during testing.

33. The testing system of claim 32, wherein the at least one condition is selected from the group consisting of device pressure, device voltage, device current, and device temperature.

Images