Automatic mode switching in volume dimensioner

Abstract

Automatic mode switching in a volumetric dimensioner. A dimensioner and method for dimensioning an object comprises capturing at least one range image of at least one field of view using a dimensioning system having a single sensor, and calculating dimension data of the range image and storing the result. Wherein, the number of captured views of the object is automatically determined based on one of three modes. If the object is a cuboid, or has no protrusions and has only one obtuse angle that does not face the viewpoint, a first mode is used, in which a single view of the object is captured. If the object includes a single obtuse angle and no protrusions, a second mode is used, in which two views of the object are captured. If the object includes protrusions and / or more than one obtuse angle, overhang, protrusion, or a combination thereof, a third mode is used, in which more than two views of the object are captured.

Description

Technical Field

[0001] The present invention relates to a system for determining the physical size of an object (ie, a dimensioning system), and more particularly to a dimensioning system and method for automatically switching modes to acquire data required for dimensioning in business. Background Art

[0002] Generally speaking, determining the dimensions of an item is often necessary as part of a logistics process in business (e.g., shipping, storage, etc.). However, physically measuring an object is time consuming and may not result in an accurate measurement. For example, in addition to human error, measurement errors may result when measuring irregularly shaped objects or when combining multiple objects into a single measurement. Therefore, dimensioning systems have been developed to automate or assist in making this measurement.

[0003] Dimensioning systems typically sense the shape / size of an object in three dimensions (3D) and then use this 3D information to calculate an estimate of the object's dimensions (e.g., volume, area, length, width, height, etc.). In addition, for irregular objects (or objects), dimensioning systems can calculate the size of a minimum bounding box (MVBB) that contains the object (or objects).

[0004] The dimensioning system can sense objects by projecting a light pattern (i.e., pattern) into the field of view. Objects within the field of view will distort the appearance of the light pattern. The dimensioning system can capture an image of the reflected light pattern and analyze the pattern distortion in the captured image to calculate the 3D data required for dimensioning.

[0005] Many dimensioners use optical devices that determine the length, width, and height of an object, which is typically a cuboid (e.g., a "normal" box with opposite sides parallel and perpendicular to adjacent sides). However, some packages are irregular and therefore a single set of dimensions may not be available, especially if one of the corners is obtuse (i.e., greater than 90 degrees). Some systems address this problem by utilizing multiple sensors, but this is expensive and multiple sensors limit the size of objects that can be measured. Another solution is to always program the device to take measurements from two or more viewpoints, but this is time consuming for a normal box or cuboid.

[0006] Therefore, there is a need for a single sensor dimensioner with an automated mechanism to detect when multiple views are needed and automatically switch modes accordingly. Summary of the invention

[0007] Thus, in one aspect, the present invention includes a method for dimensioning an object. In the method, a dimensioning system or dimensioner having a single sensor is used to capture at least one range image of an object in at least one field of view. Dimension data for at least one range image of all at least one field of view is then calculated and the results are stored. The number of views captured using the range image and the dimension data is automatically determined based on one of the following three modes: a first mode, a second mode, and a third mode. If the object is a cuboid, or has no protrusions and has only one obtuse angle that does not face the viewpoint, the first mode is automatically used, wherein the first mode captures a single view of the object. If the object includes a single obtuse angle facing the viewpoint and no protrusions, the second mode is automatically used, wherein the second mode captures two views of the object. If the object includes a protrusion and / or more than one obtuse angle, an overhang, a protrusion, or a combination thereof, the third mode is automatically used, wherein the third mode captures more than two views of the object.

[0008] One feature of the method for dimensioning may be to compute a minimum bounding box (MVBB) from range images from all automatically determined views.

[0009] In another possible embodiment, a method for dimensioning may include displaying and / or transmitting a minimum bounding box (MVBB) size for certification in a business.

[0010] In another possible embodiment, the method for dimensioning may include moving at least one of the dimensioning system and the object (e.g., the dimensioning system or the object or both) such that the field of view of the dimensioning system includes a different view of the object after one of the views is captured until a determined number of views are captured. In selected embodiments, the step of moving the dimensioning system or the object (i.e., moving the dimensioning system and / or the object) may be determined based on one or more of the following:

[0011] • Determine if there is a non-flat side of the object in the view, and if so, move the dimensioning system or the object so that the dimensioning system's field of view encompasses a different portion of the object;

[0012] • Detecting whether there is an obtuse angle facing the viewpoint, and if so, moving the dimensioning system or the object so that the dimensioning system's field of view encompasses a different portion of the object; and

[0013] • Detect if another obtuse angle and / or protrusion is present, and if so, move the dimensioning system or the object so that the dimensioning system's field of view encompasses a different portion of the object.

[0014] In another possible embodiment, the step of moving the dimensioning system or object may include generating audio and / or visual messages to guide the user to perform the movement. In selected embodiments, the audio and / or visual messages may include instructions for the user to (i) move the dimensioning system or object in a specific direction, (ii) move the dimensioning system or object at a specific speed, and / or (iii) stop moving the dimensioning system or object.

[0015] In another possible embodiment, the step of moving the dimensioning system or the object may include automatic movement of the dimensioning system or the object.

[0016] In another possible embodiment, the step of capturing a range image of the field of view using the dimensioning system may include:

[0017] • Projecting a light pattern into the field of view using a pattern projector;

[0018] • using a range camera to capture an image of the field of view, the image including the reflected light pattern; and

[0019] •Generate 3D data from images of reflected light patterns.

[0020] In another possible embodiment, the at least one range image may include 3D data sufficient to dimension the object. In selected embodiments, the 3D data sufficient to dimension may include 3D data from all necessary views of the object to calculate a minimum bounding box (MVBB). In other selected embodiments, the 3D data may be from views of the object without any gaps in the reflected light pattern.

[0021] In another possible embodiment, the dimensioning system can be handheld. For example, the dimensioning system can be incorporated into a handheld barcode scanner.

[0022] In another aspect, the present invention includes a dimensioning system including a dimensioning system having a single sensor. The dimensioning system may generally include a pattern projector, a single range camera, and a processor. The pattern projector may be configured to project a light pattern onto an object. The single range camera may be configured to (i) capture an image of the reflected light pattern in a field of view, (ii) generate 3D data from the reflected light pattern, and (iii) create a range image using the 3D data. The processor may be communicatively coupled to the pattern projector and the single range camera. The processor may be configured by software to automatically determine the number of captured views of an object based on one of the following three modes:

[0023] • If the object is a cuboid, or has no protrusions and only one obtuse corner that does not face the viewpoint, the first mode is automatically used, where the first mode captures a single view of the object;

[0024] • if the object includes a single obtuse angle facing the viewpoint and has no protrusions, automatically using a second mode, where the second mode captures two views of the object; or

[0025] • If the object includes a protrusion and / or more than one obtuse angle, overhang, protrusion, or a combination thereof, a third mode is automatically used, wherein the third mode captures more than two views of the object.

[0026] A feature of the dimensioning system may be that the processor is further configured to calculate a size of a minimum bounding box from range images from all automatically determined views.

[0027] In a possible embodiment, the dimensioning system may display and / or transmit the dimensions of a minimum bounding box (MVBB) for authentication in commerce.

[0028] In another possible embodiment, the dimensioning system may be further configured to move the dimensioning system or the object so that the field of view of the dimensioning system includes different views of the object after one of the views is captured until a determined number of views are captured. In a selected embodiment, the movement of the dimensioning system or the object is determined based on the processor being configured to:

[0029] • Determine if there is a side of the object that is not flat in the view, and if so, move the dimensioning system or the object so that the dimensioning system's field of view encompasses a different portion of the object;

[0030] • Detecting whether there is an obtuse angle facing the viewpoint, and if so, moving the dimensioning system or the object so that the dimensioning system's field of view encompasses a different portion of the object; and

[0031] • Detect if another obtuse angle and / or protrusion is present, and if so, move the dimensioning system or the object so that the dimensioning system's field of view encompasses a different portion of the object.

[0032] In another possible embodiment, the movement of the dimensioning system or object may include generating audio and / or visual messages to guide the user to perform the movement. In selected embodiments, the audio and / or visual messages may include instructions for the user to (i) move the dimensioning system or object in a particular direction, (ii) move the dimensioning system or object at a particular speed, and / or (iii) stop moving the dimensioning system or object.

[0033] In selected embodiments, the dimensioning system may be handheld. For example, the dimensioning system may be incorporated into a handheld barcode scanner.

[0034] In one possible embodiment of the above method, the 3D data sufficient for dimensioning comprises 3D data from a surface of the object without any gaps in the reflected light pattern.

[0035] In a possible embodiment of the above method, the dimensioning system is handheld.

[0036] A system includes: a pattern projector configured to project a light pattern onto an object; a single range camera configured to (i) capture an image of the reflected light pattern in a field of view, (ii) generate 3D data from the reflected light pattern, and (iii) create a range image using the 3D data; and a processor communicatively coupled to the pattern projector and the single range camera; wherein the processor is configured by software to automatically determine the number of views of the object that need to be captured based on the following modes: if the object is a cuboid, or has no protrusions and only one obtuse angle that does not face a viewpoint, automatically use a first mode, wherein the first mode captures a single view of the object; if the object includes a single obtuse angle facing a viewpoint and no protrusions, automatically use a second mode, wherein the second mode captures two views of the object; or if the object includes a protrusion and / or more than one obtuse angle, overhang, protrusion, or a combination thereof, automatically use a third mode, wherein the third mode captures more than two views of the object.

[0037] In one possible embodiment of the system, the processor is further configured to: calculate the size of the minimum bounding box from the range images from all automatically determined views; wherein the dimensioning system displays and / or transmits the size of the minimum bounding box for authentication in business.

[0038] In one possible embodiment of the system, the dimensioning system is further configured to move at least one of the dimensioning system and the object so that the field of view of the dimensioning system includes different parts of the object after capturing one of the views of the object until a determined number of views of the object are captured.

[0039] In one possible embodiment of the system, movement of at least one of the dimensioning system and the object is determined based on the processor being configured to: determine whether there is a side of the object that is not flat in the view, and if so, move at least one of the dimensioning system and the object so that the field of view of the dimensioning system includes a different portion of the object; detect whether there is an obtuse angle facing the viewpoint, and if so, move at least one of the dimensioning system and the object so that the field of view of the dimensioning system includes a different portion of the object; and detect whether there is a protrusion, and if so, move at least one of the dimensioning system and the object so that the field of view of the dimensioning system includes a different portion of the object.

[0040] In a possible embodiment of the system, the movement of at least one of the dimensioning system and the object comprises generating audio and / or visual messages to guide a user to perform the movement.

[0041] In one possible embodiment of the system, the audio and / or visual message includes instructions for the user to (i) move at least one of the dimensioning system and the object in a specific direction, (ii) move at least one of the dimensioning system and the object at a specific speed and / or (iii) stop moving the dimensioning system and / or the object.

[0042] In one possible embodiment of the system, the dimensioning system is handheld.

[0043] In one possible embodiment of the system, the dimensioning system is incorporated into a handheld barcode scanner.

[0044] The above illustrative invention content as well as other exemplary objects and / or advantages of the present invention and the manner of implementing the present invention are further explained in the following detailed description and the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A block diagram of a dimensioning system according to an embodiment of the present invention is schematically depicted.

[0046] Figure 2 The principle of sensing three dimensions using a spatially offset pattern projector and a range camera according to an embodiment of the invention is graphically depicted.

[0047] Figure 3 An implementation of a pattern projector of a dimensioning system according to an embodiment of the present invention is graphically depicted.

[0048] Figure 4 Graphically depicted is a dimensioning system and / or movement of an object according to an embodiment of the present invention.

[0049] Figure 5 Graphically depicted is a flow chart illustrating a method for dimensioning an object according to an embodiment of the present invention.

[0050] Figure 6 A two-dimensional view of an object containing obtuse angles in accordance with an embodiment of the present invention is graphically depicted.

[0051] Figure 7 Graphically depicted is a flow chart representing how one (Mode 1), two (Mode 2), or three (Mode 3) views may be automatically determined so that a minimum bounding box may be provided for use in commerce according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The present invention includes a dimensioning method and dimensioning system with a single sensor (i.e., a single image sensor and camera) that automatically switches modes for capturing a minimum number of views for dimensioning various shaped objects. The dimensioning method and dimensioner of the present disclosure are generally designed to automatically operate in one of three modes and switch between them:

[0053] • If the object is a cuboid, or has no protrusions and only one obtuse corner that does not face the viewpoint, the first mode is automatically used, where the first mode captures a single view of the object;

[0054] • if the object includes a single obtuse angle facing the viewpoint and has no protrusions, automatically using a second mode, where the second mode captures two views of the object; or

[0055] • If the object includes a protrusion and / or more than one obtuse angle, overhang, protrusion, or a combination thereof, a third mode is automatically used, wherein the third mode captures more than two views of the object.

[0056] Some advantages of using the disclosed method and system with these three automatic modes can be to quickly and accurately calculate the minimum bounding box (MVBB) for commercial use and have a similar cost to a single mode dimensioning system. Therefore, the present disclosure is designed to provide a marketable and certifiable dimensioning method and system to be used in business. The present disclosure improves on devices that are programmed to always take multiple measurements so that they are certifiable for irregularly sized objects for business. Therefore, the present disclosure can provide a lower cost but inflexible method that involves using a single algorithm that requires an operator to measure from different viewpoints (e.g., from the front and the side or from one side and the other side, etc.). In addition, the present disclosure can describe an optimized and automated method in which information about the carton or object to be dimensioned is used to determine how many views are required for commercial certification. Therefore, a device incorporating the methods and processes described herein can have approximately the same cost as a standard device with a single sensor, but can provide a unique certifiable in commercial features in the optical dimensioning market.

[0057] An exemplary dimensioning system is shown in (Fig.) 1. The dimensioning system 10 includes a pattern projector 1 configured to project a light (e.g., infrared light) pattern into a field of view 2. The light pattern typically includes light points arranged in a pattern (i.e., a point cloud). The light points can be (i) the same size or different and (ii) can be arranged in some order or pseudo-randomly. The pattern projector can create the light pattern using a light source (e.g., a laser, LED, etc.), a pattern creator (e.g., a mask, a diffractive optical element, etc.), and one or more lenses.

[0058] The dimensioning system 10 also includes a range camera 3 configured to capture an image of the projected light pattern reflected from the field of view 4 of the range camera. The field of view of the range camera 4 and the field of view of the pattern projector 2 should overlap, but may not necessarily have the same shape / size. The range camera 3 may include one or more lenses to form a true image of the field of view 4 onto an image sensor. Light filtering (e.g., an infrared filter) may also be used to help detect the reflected pattern by removing stray light and / or ambient light. An image sensor (e.g., a CMOS sensor, a CCD sensor, etc.) may be used to create a digital image of the light pattern. The range camera may also include the necessary processing (e.g., a DSP, FPGA, ASIC, etc.) to obtain 3D data from the light pattern image. As an example, and expressly not limited thereto, the range camera 3 may be based on one or more of the following: structured light, stereoscopic vision, time of flight, etc., and / or combinations thereof.

[0059] like Figure 2As shown in , the pattern projector 1 and the range camera 3 can be spatially offset (e.g., arranged stereoscopically). The spatial offset 8 takes into account changes 5 in the range of the object 6 to be detected as image offsets 7 on the image sensor of the range camera. The spatial offset 8 can be adjusted to change the image offset 7 to change the resolution at which the range difference 5 can be detected. In this way, the image offset in the point cloud pattern can be converted into 3D data of objects within the field of view of the dimensioning system.

[0060] The 3D data includes the range value for each light point in the point cloud image. In addition, the range values ​​between the light points in the point cloud image can be interpolated to create a so-called range image. A range image is a grayscale image in which each pixel value in the image corresponds to an estimated distance between the dimensioning system and a point in the field of view. A range camera can output 3D data in the form of a point cloud image or a range image.

[0061] It can be used in the processor 9 of the dimensioning system (see Figure 1 ) to analyze the range image to detect objects and determine the dimensions of the objects. In some cases, these algorithms may include a step to create a minimum bounding box (MVBB), which is a computer model of a box that encloses an object (e.g., an irregularly shaped object) or a collection of objects (e.g., multiple boxes on a pallet). In this case, the dimensioning system may return and even display the dimensions of the MVBB.

[0062] Accurate dimensioning requires a high-quality image of the reflected pattern (e.g., a point cloud image). A high-quality point cloud image is an image in which the light points in the pattern are visible on the surfaces of multiple objects. Low-quality point cloud images may be caused by a variety of situations. For example, the imaged pattern may not be visible on one or more surfaces (e.g., a surface that is blocked relative to the pattern projector) or fall outside the field of view of the pattern projector and / or range camera. In another example, the light pattern may be partially visible on the surface and / or lack sufficient pattern density (i.e., the number of visible light points on the surface). In yet another example, the illumination in the environment of the object (e.g., glare, shadows) and / or the reflectivity of the object (e.g., dark objects) may adversely affect the visibility of the light pattern.

[0063] Figure 3 A dimensioning system 10 is graphically depicted projecting a light pattern 11 onto an object 6 in a field of view 2. The depiction shows the dimensioning system 10 capturing a single view of the object 6 because the object 6 is a cuboid (mode 1).

[0064] Figure 4It is shown how movement 12 of the dimensioning system 10 and / or movement 13 of the object 6 can help capture (i.e., sense, sample, etc.) 3D data. Movement 12 and / or 13 can allow for the capture of 3D data from more portions or views of the object 6 than can be obtained using a single view. Range images can be captured and then combined to form a synthetic range image. The synthetic range image has 3D data from more points on the object. For example, all sides of the object can be sampled during the movement process to obtain 3D data from the entire object. In addition, the technique can be used to fill gaps in a pattern (i.e., missing areas in the pattern).

[0065] In one possible embodiment, the movement of the dimensioning system and / or the object is automatic and does not require user involvement. In this embodiment, the dimensioning system can be coupled to a mobile device (e.g., an actuator, a motor, etc.) that adjusts the spatial relationship between the dimensioning system and the object. In one example, an object 6 can be placed in the measurement area and the dimensioning system 10 can be moved around the object 12 to collect range images from various perspectives or views, such as Figure 4 In another example, a fixed dimensioning system is shown in Figure 4 As shown in FIG, range images are collected while the object 6 is rotating (eg, on a motorized turntable) 13. In these cases, position information may be obtained from the mobile device and may be used to help combine the range images to create 3D data.

[0066] In another possible embodiment, the movement of the dimensioning system and / or object is performed by the user. Here, messages (e.g., audio, visual, etc.) can be generated by a processor of the dimensioning system and transmitted to a user interface (e.g., a screen, indicator light, speaker, etc.). The user can follow the instructions provided by the messages to move the dimensioning system / object. The instructions can include messages to help the user know (i) how far to move the dimensioning system / object, (ii) how fast to move the dimensioning system / object, (iii) to move the dimensioning system / object to a specific location, and (iv) how long to continue moving the dimensioning system / object (e.g., when to stop moving). For example, the dimensioning system can be handheld, and the user can move the dimensioning system to change the perspective. In this case, the dimensioning system can be configured to collect tracking information (e.g., sensing its position and orientation in the environment) to help assemble the range image.

[0067] Typically, the dimensioning system can be moved in a variety of ways when capturing views and range images. However, in some cases, the movement may have certain needs to facilitate the combination. For example, the movement can be limited to movements with a constant range between the dimensioning system and the object, because changes in range affect the image size of the light pattern / object. In another example, the movement can be limited to a certain path with a specific start and end point. The path can be determined using the expected object size / shape.

[0068] The need for movement can be reduced by using simultaneous localization and mapping (SLAM). SLAM is a computer algorithm that uses an image of the environment (e.g., a range image) to update the position of an imager (e.g., a dimensioning system). When the dimensioning system is moved, for example, the SLAM algorithm can detect features (i.e., landmarks) in the captured range image and then compare these landmarks with landmarks found in the previously captured range image to update the position of the dimensioning system. This position information can be used to help assemble the range image.

[0069] Combining range images can often be accomplished using image stitching. Image stitching refers to a computer algorithm that transforms, registers, and blends multiple constituent images to form a single composite image. The image stitching algorithm can first determine a suitable mathematical model to relate the pixel coordinates of the constituent images to the pixel coordinates of a target composite image surface (e.g., a plane, cylinder, sphere, etc.). This involves transforming (e.g., warping) the images to the target composite image surface. The transformed images can then be aligned to each other (e.g., using feature detection and mapping) and fused (e.g., blending) to remove edge effects.

[0070] Although range images have pixels that represent range rather than reflected light, in most other respects they are like regular digital images. Therefore, the principles of image stitching described so far can be applied equally to range images (or point cloud images).

[0071] In one embodiment, the dimensioning system may be incorporated into a handheld barcode scanner. Typically, packages may have barcode symbols (serialized) for identification of individual items. Therefore, incorporating the present dimensioning system into a barcode scanner has a large speed advantage for looking up the barcode data online or stored in a local database to find out if the dimensions have been determined at an earlier stage of shipping. Further enhancements would be for high value items that are commonly counterfeited to compare the stored dimensions to the measured dimensions and flag the differences.

[0072] Figure 5A flow chart illustrating a method 100 for dimensioning an object is graphically depicted. The method begins by positioning 20 a dimensioning system and / or an object so that at least a portion on the object is contained within a viewpoint of the dimensioning system. The method 100 then automatically determines 110 the number of views required to dimension the object. The number of views captured using the range image and the dimension data is automatically determined based on one of three modes: a first mode, a second mode, and a third mode based on an algorithm. If the object is a cuboid, or has no protrusions and has only one obtuse angle that does not face the viewpoint, the first mode is automatically used, wherein the first mode captures a single view of the object. If the object includes a single obtuse angle facing the viewpoint and no protrusions, the second mode is automatically used, wherein the second mode captures two views of the object. If the object includes a protrusion and / or more than one obtuse angle, overhang, protrusion, or a combination thereof, the third mode is automatically used, wherein the third mode captures more than two views of the object.

[0073] Once the number of desired views is determined, the method first captures 130 a range image of the initial field of view. If the desired number of views are not captured, the dimensioning system and / or the object is then moved 160 so that another portion of the object is within the field of view and another range image is captured 130. This process of moving and capturing is repeated until the desired number of views and associated range images are captured 130.

[0074] Once the desired number of views are captured, the multiple range images may then be combined 170 to form a composite range image, and the composite range image may be used to dimension the object 190 .

[0075] In one exemplary embodiment, once dimensioning 190 is complete, the dimensions of the minimum bounding box (MVBB) may be calculated 200. In selected embodiments, these calculated dimensions of the MVBB may be displayed and / or transmitted 210.

[0076] In an exemplary embodiment, the dimensioning system may create a message 150 to direct the movement of the dimensioning system and / or the object, as previously described. In selected embodiments, the movement 150 of the dimensioning system or object may be determined based on one or more of the following:

[0077] • Determine if there is a side of the object in the view that is not flat, and if so, move the dimensioning system or the object so that the dimensioning system's field of view encompasses a different portion of the object;

[0078] • Detecting whether there is an obtuse angle facing the viewpoint, and if so, moving the dimensioning system or the object so that the dimensioning system's field of view encompasses a different portion of the object; and

[0079] • Detect if another obtuse angle and / or protrusion is present, and if so, move the dimensioning system or the object so that the dimensioning system's field of view encompasses a different portion of the object.

[0080] The present disclosure recognizes that specific features of an object to be dimensioned (i.e., a carton or a normal box) can be detected in a first view, and based on that analysis, the mode is changed, and then two or more viewpoints need to be dimensioned. The present disclosure can provide a dimensioner that can be easily moved if handheld, or alternatively, a carton placed on a static dimensioner (e.g., an auto cube) can be turned to different orientations, so that an automatic mode switching device can be used to always generate the correct dimensions, especially in countries with strict requirements for certification of dimensioning irregular objects.

[0081] Referring to Mode 1, if the carton is a cuboid or a "normal" box, then no matter from which viewpoint it is viewed, the same dimensioner results will occur (within the measurement tolerance provided by the equipment, which may be referred to as "d"). Thus, a single image will produce valid results for commercial use.

[0082] Referring to Mode 2, if the object has an obtuse angle, there is a different set of bounding box size results depending on which view the dimensioning camera has. If the obtuse angle is facing the camera, then a "shadow" or hidden area will result in making the bounding box larger than the minimum. The operator or system needs to be instructed to move the dimensioner or rearrange the object so that another view is obtained. As long as the other view is not facing the obtuse angle of the dimensioner, the result will be the minimum bounding box.

[0083] Figure 6 A two-dimensional view of an object containing an obtuse angle (angle F) according to an embodiment of the present invention is graphically depicted. If the object is viewed from the right (side b), or if the object is flipped and viewed facing away (side c pointing down) at an obtuse angle (face away), a smaller bounding box results. The minimum bounding box is the bounding box that must be reported for commercial use. Figure 6 This discussion assumes that the other end of the carton is the same and that the remaining sides are flat and perpendicular to the end. If the carton is viewed from above, as indicated by the arrow (facing side c), the dimensioner will not see the overhang and will cast a "shadow". This viewing scenario will produce a larger bounding box than if viewed from the right or if the carton is rearranged so that side "c" is downward.

[0084] Referring to mode 3, if the object has one side that is not flat, it can cause the other side to overhang and cause hidden areas. Therefore, another view will be needed to see under the overhang. Similarly, if the object has a protrusion, it can cause shadows and potentially cause a bounding box that is not minimal. Again, another view is needed to see under the protrusion. In some cases, especially if there is more than one obtuse angle, overhang or protrusion, or if there is a combination, the dimensioner must view the object from all three orthogonal sides to ensure that the minimal bounding box is captured.

[0085] In a handheld scenario, the operator may be instructed to move the dimensioner or rearrange the cartons to obtain the second and / or third images. In the case of a fixed dimensioner, the operator may be instructed to rearrange the cartons to obtain the second and / or third images, or the movement may be automatic.

[0086] Once the necessary number of views and images have been collected, the device can simply compare the values ​​and select the set of sizes that produces the smallest bounding box. If the operator has not collected the necessary images, the device must not provide any sizes (e.g., issue a static failure notification).

[0087] Reference now Figure 7 , an overview of one embodiment of automatic mode switching in a volume dimensioner is shown as a flow chart. The flow chart shows how one (mode 1), two (mode 2), or three (mode 3) viewpoints can be automatically determined so that the dimensions of the MVBB can always be provided. An overview of the present disclosure is that the dimensioner automatically determines whether it has a safe view of an object and can guarantee that the smallest bounding box is reported. Ultimately, it may be necessary for a customer to need three orthogonal views so that the customer will not be overcharged. Therefore, the dimensioner can be certified for use in business.

[0088] In the specification and / or drawings, typical embodiments of the present invention have been disclosed. The present invention is not limited to such exemplary embodiments. The use of the term "and / or" includes any and all combinations of one or more of the associated listed items. The drawings are schematically identified and therefore not necessarily drawn to scale. Unless otherwise specified, specific terms have been used in a general and descriptive sense and are not for limiting purposes.

Claims

1. A dimensioning system, comprising: a range camera configured to capture a first light pattern image projected onto the object from a first view; generating a first range image based on the first light pattern image in response to determining that the object has flat sides and an obtuse angle that does not face a viewpoint of the dimensioning system; as well as a processor communicatively coupled to the pattern projector and the range camera, wherein the processor is configured by software to: calculating a first set of dimensions of the object based on the first range image; causing the range camera to capture a second light pattern image of the object from a second view; generating a second range image based on the second light pattern image, in response to determining based on the first set of dimensions that the object has an uneven side and an obtuse angle facing a viewpoint of the dimensioning system, wherein the second view is different from the first view; calculating a second set of dimensions of the object based on the second range image; and A minimum bounding box of the object defined by the projected light pattern onto the object is determined based on at least the first set of sizes and the second set of sizes.

2. The dimensioning system of claim 1 , wherein in response to determining the minimum bounding box of the object, the processor is configured by software to further: calculating a first bounding box of the object based on the first set of sizes; calculating a second bounding box of the object based on the second set of sizes; A smaller bounding box is determined from the first bounding box and the second bounding box as a minimum bounding box.

3. The dimensioning system of claim 1 , wherein in response to determining the minimum bounding box of the object, the processor is configured by software to further: generating a composite image based on the first range image and the second range image; and Determine the minimum bounding box of the object based on the composite image.

4. The dimensioning system of claim 1 , wherein the processor is configured by software to: determining, based on the first range image, that the object includes at least one of: (a) a protrusion, (b) an uneven side, or (c) two or more obtuse angles; causing the range camera to capture a third light pattern image of the object from a third view and generate a third range image based on the third light pattern image, wherein the third view is different from the first view and the second view; and calculating a third set of dimensions of the object based on the third range image; Determining a minimum bounding box of the object is based on at least a first set of sizes, a second set of sizes, and a third set of sizes.

5. The dimensioning system of claim 4, wherein in response to determining the minimum bounding box of the object, the processor is configured by software to further: calculating a first bounding box of the object based on the first set of sizes; calculating a second bounding box of the object based on the second set of sizes; calculating a third bounding box for the object based on the third set of sizes; and A minimum bounding box is determined from the first bounding box, the second bounding box, and the third bounding box as a minimum bounding box.

6. The dimensioning system of claim 4, wherein in response to determining the minimum bounding box of the object, the processor is configured by software to further: generating a composite image based on the first range image, the second range image, and the third range image; and Determine the minimum bounding box of the object based on the composite image.

7. The dimensioning system of claim 1 , wherein the object comprises a barcode symbol, wherein the dimensioning system further comprises a barcode scanner, and wherein the processor is configured by software to further: Enables a barcode scanner to scan a barcode symbol; retrieving a set of historical dimensions of the object from a database based on the bar code symbology; and The minimum bounding box is compared to a set of historical sizes of the object.

8. The dimensioning system of claim 7, wherein the processor is configured by software to further: determining that the minimum bounding box is inconsistent with the set of historical sizes of the object; and Marks the object as a fake.

9. The dimensioning system of claim 1, wherein the range camera utilizes at least one of: structured light, stereo vision, or time of flight.

10. The dimensioning system of claim 1, wherein the object is on a motorized turntable, and wherein in response to causing the range camera to capture the second light pattern image, the processor is configured by software to cause the motorized turntable to rotate.

Images