WO2014150739A1 - Alignement de masque virtuel pour analyse d'ajustement - Google Patents

Alignement de masque virtuel pour analyse d'ajustement Download PDF

Info

Publication number
WO2014150739A1
WO2014150739A1 PCT/US2014/024098 US2014024098W WO2014150739A1 WO 2014150739 A1 WO2014150739 A1 WO 2014150739A1 US 2014024098 W US2014024098 W US 2014024098W WO 2014150739 A1 WO2014150739 A1 WO 2014150739A1
Authority
WO
WIPO (PCT)
Prior art keywords
mask
ppe
facial
virtual
fit
Prior art date
Application number
PCT/US2014/024098
Other languages
English (en)
Inventor
Xiaoli Wang
Henry Chen
Yi Sun
Paul Derby
Hari Thiruvengada
Original Assignee
Honeywell International Inc.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from US13/839,056 external-priority patent/US9361411B2/en
Priority claimed from US13/839,186 external-priority patent/US20140278320A1/en
Application filed by Honeywell International Inc. filed Critical Honeywell International Inc.
Priority to US14/906,832 priority Critical patent/US10424113B2/en
Publication of WO2014150739A1 publication Critical patent/WO2014150739A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T19/00—Manipulating 3D models or images for computer graphics
    • G06T19/20—Editing of 3D images, e.g. changing shapes or colours, aligning objects or positioning parts
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
    • G06V10/757—Matching configurations of points or features
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2219/00—Indexing scheme for manipulating 3D models or images for computer graphics
    • G06T2219/20—Indexing scheme for editing of 3D models
    • G06T2219/2021—Shape modification

Definitions

  • Various embodiments relate generally to personal protective equipment and more specifically to a system and method for predicting an optimal fit of a respirator to a facial area.
  • PPE Personal Protective Equipment
  • respirators are widely used in a variety of different applications.
  • many workplaces that subject an employee to hazardous atmospheric conditions require the employee to wear respiratory protection for several hours per day.
  • respiratory protection requires a proper seal upon a facial area of the user. A poor seal and thus poor fit may result in leakage and the possibility of the inhalation of contaminants.
  • an employee may not be able to find a respirator having a suitable fit. For example, the employee may not be given adequate time to try on different respirators or the employee may not be given an adequate variety of respirator samples to try.
  • Apparatus and associated methods relate to fitting a virtual mask to a virtual face by first fitting a chin region of the virtual mask to the virtual face, then determining a virtual mask angle that maintains the fitted chin region while simultaneously fitting a nose- bridge region of the virtual mask to the virtual face, and then calculating a fit-quality metric corresponding to the fitted position.
  • the fitted chin region may include the high curvature menton region of the chin.
  • a virtual mask may be virtually pressed toward the virtual face using a predetermined force corresponding to a force of a mask securing device of a real mask corresponding to the virtual mask
  • the fitting of a virtual mask to a virtual face may advantageously yield a mask's fit quality in a brief amount of time.
  • fit tests may be performed to ensure that the mask seals properly to a person's face. Fit tests may be time consuming, as the person first may need to select a mask for testing, and then don the selected mask. After donning the selected mask, the person may subject him/herself to a qualitative test. The person may then stand in a testing chamber in which the ambient may be exposed to chemicals that the user can detect by smell or taste, for example. If the user detects the chemical introduced into the ambient, the mask seal may be determined to be inadequate. Such a qualitative test may take tens of minutes to complete. And the results of the test are not precise as to the quality of the fit. For example, one may not be able to determine if the mask fit has a small seal leak or a large seal leak.
  • the person may then undergo a quantitative test.
  • the mask wearing person may have a machine connected to a mask portal via a tube.
  • the machine may then monitor the quality of the exhalations from the mask portal.
  • Measurable chemicals may be introduced into the testing chamber. If the chemicals are detected in the exhalation chemistry, the machine may measure the concentration of the detected chemical. The machine may then determine a magnitude of the mask seal leak.
  • This test may take additional tens of minutes to perform. After performing the above described tests, the person often may have to repeat the testing process wearing another mask selected for testing. Such repetitions can be very time consuming and/or expensive.
  • some embodiments may facilitate the virtual fitting of many masks to a user in a brief amount of time.
  • the time and cost of performing qualitative and quantitative testing of masks that are unlikely to fit well may be eliminated.
  • an ability to suggest a mask having a good likelihood to provide a proper seal may result in more comfortable mask assignments. Such comfort may translate into improved worker productivity and/or increased mask use.
  • a database of users' 3D virtual faces may be used to direct inventory decisions.
  • the database of 3D virtual faces may direct future mask development activities. Safety masks having improved fit for a variety of faces may result from using such a database.
  • Apparatus and associated methods may relate to determining a fit-quality metric for a mask/face combination based upon a calculated dead-space volume between a virtual mask and a virtual face virtually aligned so as to create an integrity seal
  • an interactive virtual fitting system may receive a three-dimensional (3D) virtual face associated with a person.
  • the system may retrieve 3D models of various respirators selected by user determined criteria.
  • the system may then compute a fit-quality metric for each of the retrieved 3D models.
  • the potential wearer may then be presented with the metrics for review.
  • the potential wearer may select a respirator based upon these computed metrics.
  • a virtual fitting of many respirators may advantageously reduce the time needed for selecting a properly fitting respirator while simultaneously ensuring that the selected respirator may be comfortable and well fitting.
  • Some embodiments may reduce the time needed for a person to be fit to a
  • the person to be fit may need not be present at the location where the virtual fitting may be computed.
  • the person to be fit may need not be present at a storage facility for PPE devices.
  • Sample fitting devices need not be purchased. The elimination of sample fitting devices may preclude the need for cleaning such devices between fittings. Inventories of PPE devices may be reduced by the elimination of sample fitting devices. [001 1] Potential wearer's of PPE devices may have more devices virtually fit than they would physically try on. This increased fitting count may translate into improved matching of PPE device to a wearer. Wearer's may find more comfort in their selected PPE devices. And PPE devices may properly fit a wearer. This proper fit may translate into improved protection against harm.
  • FIG. 1 depicts an exemplary PPE virtual fitting station.
  • FIG. 2 depicts an exemplary mask face alignment along the mid-sagittal plane cross-section.
  • FIG. 3 depicts a side view of a 3D captured human face and the determination of facial profile along the mid-sagittal plane.
  • FIG. 4 depicts an exemplary breathing mask cross-sectioned along the mid- sagittal plane.
  • FIG. 5 depicts the surfaces of an exemplary breathing mask along the mid- sagittal plane of the mask.
  • FIGS. 6A-6C depict the facial contacting menton region of an exemplary breathing mask and the corresponding menton region of a virtual face along the mid-sagittal plane.
  • FIG. 7 depicts the juxtaposition of a virtual mask on a virtual face, aligned to the best fit of the menton region.
  • FIG. 8 depicts a schematic of a virtual mask being rotated into an optimal fit position on a user's virtual face.
  • FIG. 9 depicts an exemplary breathing mask with exemplary landmark points identified around a sealing periphery.
  • FIG. 10 depicts an exemplary method for predicting the fit of a virtual mask to a virtual face.
  • FIG. 1 1 depicts an exemplary method for aligning a menton region of a virtual mask to a menton region of a virtual face.
  • FIG. 12 depicts an exemplary method for determining a rotation angle between a virtual mask to a virtual face.
  • FIG. 13 depicts an exemplary method for pressing a virtual mask into a virtual face.
  • FIG. 14 depicts a block diagram of an exemplary virtual fitting system.
  • FIG. 15 depicts an exemplary GUI screenshot during a three-dimensional facial acquisition.
  • FIG. 16 depicts an exemplary depiction of the facial model compared with a user's face.
  • FIG. 17 depicts an exemplary GUI screenshot of a PPE device presentation to a user.
  • FIG. 18 depicts an exemplary GUI screenshot of the presentation of a fitting metric resulting from the computation of a virtual fit.
  • FIG. 19 depicts an exemplary GUI screenshot of the presentation of a facial sealing location of a PPE device as computed during a virtual fit.
  • FIG. 20 depicts an exemplary GUI screenshot of a sealing metric of a PPE device as computed during a virtual fit.
  • FIG. 21 depicts an exemplary GUI screenshot of a dead-space metric of a
  • PPE device as computed during a virtual fit.
  • FIG. 22 depicts a block diagram of an exemplary virtual fitting system.
  • FIG. 23 depicts a flow chart of an exemplary virtual fitting system.
  • FIG. 24 depicts an overview of an exemplary respirator selection system.
  • FIG. 25 depicts a flowchart of an exemplary PPE selection system.
  • FIG. 26 depicts an overview of an exemplary PPE selection system.
  • FIG. 27 depicts a flowchart of an exemplary process of a static modeling module.
  • FIG. 28A depicts a flowchart of an exemplary dynamic modeling module using actual body movement.
  • FIG. 28B depicts an exemplary graphical view of the dynamic modeling module of Figure 28A.
  • FIG. 29A depicts a flowchart of an exemplary dynamic modeling module using simulated body movement.
  • FIG. 29B depicts a graphical view of the exemplary dynamic modeling module of FIG. 29A.
  • FIG. 30 depicts an overview of another exemplary PPE selection system.
  • FIG. 31 depicts a flowchart of an exemplary PPE selection system.
  • FIG. 32A depicts a flowchart of another exemplary PPE selection system.
  • FIG. 32B depicts an exemplary center part on a ROI as defined with reference to FIG. 32A.
  • FIG. 33 depicts a flowchart of an exemplary deformation process.
  • FIG. 34 depicts a graphical representation of an exemplary color-coded display of a PPE fit.
  • FIG. 35 depicts a flowchart of an exemplary color-coded result generator.
  • FIGS. 1-13 An exemplary method of virtually aligning a virtual mask to a virtual face will be described with reference to FIGS. 1-13.
  • FIGS 14-22 an exemplary virtual mask- fitting system that implements some of the virtual alignment techniques described in FIGS 1- 13.
  • the discussion of alignment techniques begins with a description of an exemplary virtual alignment system, with reference to FIGs. 1 and 10.
  • an exemplary chin region alignment technique will be described with reference to FIGS. 2-6C and 11.
  • FIGS. 2-6C An exemplary nose-bridge fitting technique, with reference to FIGS. 7-8 and 12.
  • An exemplary virtual mask-tightening procedure will be described with reference to FIGS. 9 and 13.
  • FIGS. 14-21 The discussion of a virtual mask- fitting system begins, with reference to FIGS. 14-21, by describing an exemplary PPE selection by a user using a virtual fitting station. Then, with reference to FIG. 22, a description of exemplary system components for a virtual fitting station will be detailed. Then, with reference to FIG. 23, an exemplary method of PPE selection using a virtual fitting station will be described.
  • FIG. 1 depicts an exemplary PPE virtual fitting station.
  • a user 100 is seated at a virtual fitting station 110.
  • the virtual fitting station includes a 3D facial scanner 105, a display device 140 and a mask fitting engion 170.
  • the mask fitting engine 170 is depicted having a processor 165 which may execute instructions received from program memory locations 175. The received program instructions may cause the processor 165 to perform operations related to obtaining data representative of a 3D facial model representative of a user's face, aligning the facial model to a 3D mask model, and calculating a fit-quality metric characterizing a quality of the mask model fit to the facial model.
  • Such processor operations and/or other operations be repeated on other 3D mask models representative of different models and/or makes of masks.
  • Such processor operations and/or other operations may be repeated on other facial models representative of other users' faces.
  • These operations may advantageously return fit-quality metrics for one or more user/mask combinations in a short time. These fit-quality metrics may select real mask models, which may be tested further for fit integrity. Such further fit integrity testing may be avoide for user/mask combinations associated with a poor virtual fit-quality metric.
  • a user 100 is operating a 3D facial scanner 105 at a virtual fitting station 110.
  • the 3D facial scanner 105 has scanned a user's face 115 creating facial elevation data 120 and corresponding facial position data 125.
  • a virtual fitting engine 130 may create a virtual face 135 based upon the scanned facial elevation data 120 and the corresponding facial position data 125.
  • the virtual fitting engine 130 may send image data corresponding to the virtual face 135 to a display device 140.
  • the virtual fitting engine 130 may retrieve a virtual mask 145 from a mask database 150.
  • the virtual mask 145 may include a virtual facial-contacting surface 155 configured to contact a virtual face 135 circumscribing a nose and mouth region of a virtual face 135.
  • the virtual fitting engine 130 may virtually align the virtual mask 145 to the virtual face 135 near a high curvature location of a chin region.
  • the virtual fitting engine 130 may virtually rotate the virtual mask 145 about a chin region rotation point to bring a nose-bride region of the virtual mask 145 into virtual proximity with a nose-bridge region of the virtual face 135.
  • the virtual mask 145 may be virtually pressed against the virtual face 135 to simulate a force of a real mask attachment device when securing a real mask to a real face.
  • the virtual fitting engine 130 may then calculate a fit-quality metric 160 based upon a relationship of the virtual facial-contacting surface 150 to adjacent facial elevation data 120 when the virtual mask 145 has been aligned as described above.
  • the fit-quality metric 160 may be calculated by evaluating the proximity of a virtual facial-contacting surface 155 along a path circumscribing the nose and mouth region of the virtual face 135.
  • the calculated fit-metric 160 may be associated with a mask-user combination.
  • the virtual fitting engine 130 may send a signal representative of the fit-quality metric 160 for display to the display device 140. After an initial scan of the user's face 1 15, many different virtual masks may be virtually fit to the virtual face 135 associated with the user 100 perhaps without need of the users continued participation. In this way, one or more well-fitting masks may be optimally selected from the mask database 150 for trial by the user 100.
  • FIG. 2 depicts an exemplary mask/face alignment along a mid-sagittal plane cross-section.
  • an exemplary virtual mask 200 and an exemplary virtual face 205 are shown in cross-section along a mid-sagittal plane (the plane of the drawing sheet).
  • a menton region 210 and a sellion region 215 of the virtual mask are identified.
  • the virtual face 205 also has a menton region 220 and a sellion region 225.
  • a coordinate system depicted in the figure aligns a Y-axis 220 in the vertical direction and a Z- axis 225 in the horizontal direction within the mid-sagittal plane.
  • An X-axis may extend perpendicular to the drawing sheet.
  • the virtual mask 200 may be fit to the virtual face 205 first at the menton region 210.
  • the menton region 220 of the virtual face 205 may have high curvature in the Y-Z plane at a forward chin location.
  • the virtual mask 200 may have a complementary high-curvature in the menton region 210 for receiving the high-curvature of the forward chin location. This high-curvature region may provide an initial fitting relationship between the virtual mask 200 and the virtual face 205.
  • the virtual mask 200 may then be rotated about the fitted menton region 210 to optimally fit the sellion region 215 of the virtual mask 200 to the sellion region 225 of the virtual face 205.
  • the sellion region 215 may first be fit and then the mask rotated to fit the menton region 210.
  • a small fraction of the virtual contact points of the mask/face interface may be considered. This, in turn may result in an initial mask/face alignment being quickly determined.
  • FIG. 3 depicts a side view of a 3D captured human face and the determination of facial profile along the mid-sagittal plane.
  • a captured 3D human face 300 is displayed from a side perspective so that a mid-sagittal plane profile 305 may be clearly depicted.
  • the mid-sagittal plane has a Z-axis 310 and a Y-axis 325.
  • the maximal projection in the Z-direction 310 of the image may be used to identify a nose tip 315 of the captured 3D human face 300.
  • a location of a menton region 320 may then be obtained using the relative Y and Z positions of the chin from the nose tip 315.
  • the curvature of the mid-sagittal plane profile 305 may be used to determine the location of the menton region 320. Locating the menton region 320 from the captured 3D human face 300 may provide an initial location for fitting a virtual mask, for example.
  • the mid-sagittal plane may be determined by finding a line of facial symmetry, for example. The determined mid-sagittal plane may be used to center a virtual mask upon a virtual face. After centering a virtual mask upon a virtual face, a vertical (Y-axis) location may be obtained for aligning a virtual mask to a virtual face. Where a face has little curvature along this vertical dimension, the face may present few locating features.
  • locating features may be presented. If a virtual mask is designed to engage such a large curvature locating feature (e.g. chin or nose), for example, locating these features may assist in finding a good initial position for mask alignment.
  • a virtual mask is designed to engage such a large curvature locating feature (e.g. chin or nose), for example, locating these features may assist in finding a good initial position for mask alignment.
  • FIG. 4 depicts an exemplary breathing mask cross-sectioned along the mid- sagittal plane.
  • an exemplary virtual mask 400 is shown from a rear perspective view.
  • the exemplary virtual mask 400 has been cross-sectioned along the mid- sagittal plane so that a mid-sagittal elevation profile 405 may be seen.
  • a facial-contacting surface 410 of the depicted virtual mask 400 is shown.
  • a real contacting surface of a corresponding real mask may be made of a soft and/or deformable material. Such a deformable material may facilitate conformity of the mask to a user's face when the mask is pressed into the face by a mask securing device.
  • a facial contacting region of the mid-sagittal elevation profile 405 may be used for an initial locating of the virtual mask to the virtual face, for example.
  • the menton region of the facial contacting surface 410 may be used to initially locate the virtual mask to the virtual face.
  • the facial contacting surface 410 may present a curvature region that is substantially complementary to a facial curvature feature.
  • Masks may be designed to engage a facial feature having a large curvature. Some masks may have malleable engagement surfaces that may deform when pressed into a user's face.
  • a virtual mask 400 may include indicia of malleability.
  • the malleability of a facial-engagement surface 410 may be a function of location along the facial-engagement surface 410. This malleability function may be used in determining a pressed position of a mask against a user's face, for example. Pressing a mask against a wearer's face may minimize or eliminate air gaps in the perimeter of the engagement-surface/user' s-face interface.
  • FIG. 5 depicts the surfaces of an exemplary breathing mask along the mid- sagittal plane of the mask.
  • the virtual mask data 500 plotted in a Y- Z coordinate system.
  • the virtual mask data 500 is the mid-sagittal profile of a virtual mask.
  • the origin of the Y-Z coordinate system centers the mid-sagittal profile data 500 in the coordinate system.
  • the mid-sagittal profile 500 includes a menton contacting surface 505 and a nose-bridge contacting surface 510.
  • the menton contacting surface 505 is depicted with a curvature in the Y-Z plane.
  • the menton contacting surface may be obtained by first extracting points in the third quadrant 515 of the depicted coordinate system, for example.
  • the mid-sagittal profile data 500 in the third quadrant 505 may include both a menton contacting surface 505 and non-contacting surfaces 520.
  • This menton contacting surface 505 may be extracted by selecting a subset of the third quadrant data having the most negative Z-coordinates, for example. Then those data that are monotically increasing as their Y-coordinates increase (concave-up curvature) may be selected from this most negative Z-coordinate subset.
  • This selected data may be used to fit the virtual mask to a virtual face at a menton region, for example.
  • the virtual mask may identify the menton contacting surface 505 as part of the model.
  • the selection of the menton contacting surface may need to be performed only once when a virtual mask model is being created. The selection of the nose-bridge contacting surface may be found in similar manners.
  • FIGS. 6A-6C depict the facial contacting menton region of an exemplary breathing mask and the corresponding menton region of a virtual face along the mid-sagittal plane.
  • third quadrant menton region data 600 is displayed.
  • the third quadrant menton region data 600 includes a facial contacting surface 605 and non-contacting surfaces 610.
  • the facial contacting surface 605 has been aligned to a mid-sagittal facial profile 615 of a person.
  • a least-squared fitting method may be used to determine the optimal fit location of the facial contacting surface 605 to the mid-sagittal facial profile 615.
  • a sliding window analysis may be used to determine an optimal fit location.
  • interpolation between adjacent data points may be used to increase the density of the data points to be fit.
  • interpolation may be used to provide a common coordinate system for data comparison and/or fit.
  • the optimal fit may be determined by aligning the points of maximum curvature of the virtual mask to that of the virtual face, for example.
  • elasticity values may be assigned to each data point of the menton region data facial contacting surface 605. The elasticity values may be a function of the location of each data point, for example. In some embodiments, the elasticity values may be used in determining the optimal fit location. For example, a weighting factor corresponding to the elasticity may be assigned to each data point.
  • a highly elastic data point may be assigned a low weighting value. This may permit the data point to fit more poorly to the face than a data point with a high weighting value corresponding to a low elasticity.
  • FIG. 6C depicts a close-up view of the facial contacting surface 600 positioned in the optimal fit location with respect to the mid-sagittal facial profile 615 of the person.
  • FIG. 7 depicts the juxtaposition of a virtual mask on a virtual face, aligned to the best fit of the menton region.
  • a virtual mask 700 has been translated to a virtual face 705.
  • the translation may have aligned a previously determined optimal fit position of the menton region 710 of the virtual mask 700 to that of virtual face 705.
  • a sellion region 715 of the virtual mask 700 may be poorly aligned relative to a sellion region 715 of the virtual face 705.
  • the virtual mask 705 may be rotated such that the sellion region 715 of the virtual mask 700 is pressed into the virtual face 705. Such a rotational position would likely be uncomfortable or even impossible, even with consideration of a mask's elasticity.
  • the virtual mask 700 may need rotation away from the sellion region 715 of the virtual face 705 to obtain a better quality fit.
  • the virtual face 705 may be rotated to accommodate any rotational mismatch between a virtual mask 700 and a virtual face 705.
  • FIG. 8 depicts a schematic of a virtual mask being rotated into an optimal fit position on a user's virtual face.
  • a user's virtual face 800 has a virtual mask 805 superimposed onto it at an initial alignment position.
  • a chin region 810 has been fit, but a nose-bridge region 815 requires fit modification.
  • the angle between the virtual mask 805 and the virtual face 800 may require modification.
  • a determination of an optimal angle may include a determination of an optimal nose-bridge mask/face alignment, for example.
  • a point of rotation may be selected near the chin region 810, for example.
  • the point of rotation may be selected at the central point of a menton contacting surface, for example.
  • a point near a forward chin-bone location of the virtual face 800 may be selected as the point of rotation.
  • a least squares rotation operation may be performed.
  • the optimum rotation may be performed using weighted data corresponding to elasticities of the mask at various locations along a facial-contacting surface.
  • a fit metric corresponding to each data point may be based upon the difference in a Z-direction between the virtual mask 800 and the virtual face 805.
  • a fit metric corresponding to each data point may be based upon the distance between the virtual mask 800 and the virtual face 805 normal to a surface of the virtual mask 800 and/or the virtual face 805.
  • FIG. 9 depicts an exemplary breathing mask with exemplary landmark points identified around a sealing periphery.
  • an exemplary virtual mask 900 has a contacting surface 905 configured to peripherally seal the virtual mask 900 to a virtual face around a nose and mouth region. After fitting a virtual mask 900 to a virtual face, some embodiments press the virtual mask 900 into the virtual face. This pressing operation may simulate a real force imparted by a real mask onto a real face by a real mask securing device.
  • an elastic band may serve as a mask securing device.
  • a tightening strap may serve as a mask securing device.
  • a real mask securing device may force a mask into a face to provide an integrity seal between the mask and the face of a wearer.
  • a functional relationship may be defined between a deformability parameter and a location of the contacting surface 905.
  • each location of the contacting surface may have a corresponding deformability value.
  • the virtual mask 900 may be translated in the direction of a virtual face in response to a virtual force. The translation may proceed until the virtual force required for such a translation exceeds a predetermined threshold, for example.
  • the virtual force may be calculated by integrating the force required for deformation around the periphery of the contacting surface 905 of the virtual mask 900.
  • the initial position of the mask may contact the user at three distinct points before the translation begins.
  • the mask may be incrementally translated in the facial direction. After each translation, an integrated force may be calculated around the sealing periphery.
  • the translation may be terminated when the calculated force exceeds a predetermined threshold. In some embodiments the translation may be terminated when a maximum force at any discrete location exceeds a predetermined threshold.
  • the mask may be further rotated after a force has been calculated. For example, an integrated force on an upper region of the mask may be much greater than an integrated force on a lower region of the mask. To balance the upper and the lower force, the mask may be rotated, for example. Further translation may again be resumed after the mask forces have been equilibrated, for example.
  • an integration may be approximated by a summation of discrete landmark points around a mask periphery.
  • Many landmark points may be evaluated, in some embodiments.
  • landmark points may be selected around a sealing surface that may circumferentially seal around a user's mouth and nose.
  • Each landmark point may be assigned a toleration value, a component of which may represent a flexion value of the mask at that landmark point location, and a component of which may represent a reasonable tolerance of tissue deformation at a facial position associated with the landmark point location.
  • Various numerical means of determining an optimal rotation may be used. For example, such techniques as least-squared fit methods may be used in determining the optimal rotation of the virtual mask. In some examples, weighted regression techniques may be used to determine the optimal rotation.
  • FIG. 10 depicts an exemplary method for predicting the fit of a virtual mask to a virtual face.
  • an exemplary virtual fitting method 1000 is described from a vantage point of the processor 165 depicted in FIG. 1.
  • the virtual fitting method 1000 begins with the processor 165 receiving a 3D virtual face of a person 1005.
  • the 3D virtual face may be obtained by a 3D scanner 105, for example, and then communicated to the processor 165.
  • the 3D virtual face may have been previously obtained and retrieved by the processor 165 from a storage device.
  • the processor 165 retrieves a virtual mask from a database of virtual masks 1010. Then the processor 165 determines an optimal chin alignment between the virtual mask and the virtual face 1015.
  • the processor 165 determines an optimal nose-bridge fit by rotating the virtual mask toward or away from the virtual face about the optimal chin location 1020.
  • the processor 165 then translates the virtual mask, as aligned at the chin and rotated to the nose-bridge, into the virtual face 1025. This translation may simulate the translation of a real mask into a real face in response to an applied force.
  • a mask securing device may produce the force that secures a mask to a face, for example.
  • the processor 165 may then calculate a fit-quality metric and associate the fit-quality metric with the virtual face-mask combination 1030.
  • the processor 165 may test whether this fit-quality metric is the best one associated with the virtual face 1035.
  • the fit-quality metric is assigned as the best fit-quality metric 1040. If, however, the fit-quality metric is not the best fit-quality metric associated with the virtual face, the previous best fit-quality metric remains. Regardless of the outcome of the best fit-quality metric test, the processor then determines whether all of the virtual masks have been evaluated for this virtual face 1045. If not, the processor returns to step 1010. If, however, all of the virtual masks have been evaluated for this virtual face, then the list of fit metrics is displayed in descending order of fit quality and the method ends 1050.
  • FIG. 11 depicts an exemplary method for identifying a menton region of a virtual face.
  • an exemplary menton locating method 1100 is described from a vantage point of the processor 165 depicted in FIG. 1.
  • the exemplary menton locating method 1100 may be performed as part of the fit menton region step 1015 of the fit prediction method 1000, for example.
  • the menton locating method 1 100 begins by identifying facial features in a virtual face 1 105. For example, eyes, nose and a mouth may be identified by characteristics that are unique to each of these features.
  • a nose-tip for example, may be located by locating the highest Z-elevation location in the virtual face.
  • the processor 165 calculates a mid-sagittal plane of symmetry 11 10.
  • the processor may, for example, convolve a mirror image of the virtual face with the non-mirrored virtual face.
  • the translated location where the mirrored image best matches the non-mirrored image may then be used to find a line of symmetry.
  • a line of symmetry may be determined by finding the image locations where the mirrored image aligns with the non-mirrored image.
  • the processor 165 initializes an index N 11 15.
  • the index may correspond to starting location on the line of symmetry from which the processor will begin its search for the menton region of the virtual face.
  • the starting location may, for example, begin just beneath the location of the nose-tip.
  • the processor then retrieves the first three facial elevation points, Z N _i, Z and ZN+I , along the line of symmetry going from the starting point down toward the lower face 1020.
  • a Y-Z curvature value may be calculated from these the three facial elevation points. For example, the following equation may be used as a relative curvature metric:
  • C is a measure of the curvature.
  • the processor compares the calculated value of curvature with zero 1 125. If the calculated curvature is less than zero, then the
  • FIG. 12 depicts an exemplary method for identifying a menton region of a virtual face.
  • an exemplary angular alignment method 1200 is described from a vantage point of the processor 165 depicted in FIG. 1.
  • the exemplary angular alignment method 1200 may be performed as part of the rotate mask step 1020 of the fit prediction method 1000, for example.
  • the angular mask alignment method 1200 begins by the processor 165 selecting a pivot point about which the mask may be rotated 1205.
  • the location of the pivot point may be selected to minimize the change in the fit quality of the menton region, for example.
  • the pivot point may be located substantially at a central location of the fit menton region.
  • the processor 165 then receives mid-sagittal facial nose-bridge elevation data of a virtual face 1210.
  • the processor 165 then retrieves mid-sagittal mask nose-bridge elevation data of a virtual mask 1215.
  • the processor 165 then calculates a rotational angle for aligning the received facial nose-bridge elevation data to the retrieved nose-bridge mask elevation data 1220.
  • the processor 165 then virtually rotates the virtual mask.
  • the processor 165 may calculate new coordinates for each of the virtual mask surfaces based upon the calculated rotational angle for aligning the nose-bridge regions of the virtual mask and the virtual face. These new coordinates may include those for the menton region, in some examples.
  • the processor 165 determines if the menton region requires being refit 1230.
  • the processor 165 may assess whether the rotated facial contacting surface of the virtual mask remains well-fitted to the menton region of the virtual face after rotation. If the processor 165 determines that refitting the menton region is required, the processor 165 will again fit the menton region of the rotated virtual mask to the menton region of the virtual face 1235. The method will then return to step 1205 and repeat the rotation of the mask by again selecting a pivot point. If, however, at step 1230, the processor determines that the menton region does not require refitting, the method ends.
  • FIG. 13 depicts an exemplary method for tightening a virtual mask.
  • an exemplary mask tightening method 1300 is described from a vantage point of the processor 165 depicted in FIG. 1.
  • the exemplary mask tightening method 1300 may be performed as part of the translate mask step 1025 of the fit prediction method 1000, for example.
  • the mask tightening method 1300 begins with the processor 165 retrieving elasticity data associated with a facial-contacting surface of a virtual mask 1305.
  • the elasticity data may correspond to a physical parameter of the mask.
  • the elasticity data may be a function of the facial-contacting surface location.
  • the elasticity data may correspond to a facial skin thickness that varies as a function of facial location.
  • the elasticity data represents a combination of physical mask materials and physiological facial statistics.
  • the processor 165 calculates an integrated force along the sealing interface between the virtual mask and the virtual face 1310. For example, the processor 165 may calculate an integrated force at the mask position aligned at a menton region and rotated to align a nose-bridge region. The processor 165 then determines if the calculated force exceeds a predetermined threshold 1315. If the forced does exceed the predetermined threshold, then the processor translates the virtual mask toward the virtual face by an incremental amount 1320.
  • the direction of translation may be in a fixed Z direction, for example. In some embodiments, the direction of translation may be in a substantially normal direction to the general plane of the sealing interface.
  • the direction of translation may be determined by the relative forces calculated around the periphery of the sealing interface.
  • the processor then returns to step 1310 and recalculates the integrated force in the translated mask position. If, however, at step 1315, the calculated force exceeds the predetermined threshold, then the method ends.
  • FIG. 14 depicts a block diagram of an exemplary virtual fitting system.
  • an exemplary virtual fitting system 1400 includes a control system 1405 and a 3D facial scanner 1410.
  • the control system 1405 is in communication with a virtual fitting engine 1415 and an inventory modue 1420.
  • the control system 1405 may be running a GUI program retrieved from program memory 1425.
  • the GUI program may contain instructions that, when exectuted, may coordinate the acquisition of a 3D face, the virtual fitting of a 3D mask to the 3D face, ordering masks from venders, as well as many other PPE related activities, for example.
  • the fitting engine 1415 may perform various alignment operations to align a virtual mask to a virtual face.
  • the fitting engine 1415 may have a rotaion module 1425.
  • the fitting engine 1415 may have a translation module. Some embodiments may have rotation and/or translation modules that operate in more than one dimension. For example, a rotation may be performed about an X- axis, a Y-axis or a Z-axis, or a combination thereof.
  • the fitting engine 1415 may keep statistics for use in future product development activities.
  • the inventory module 1420 may have databases of masks 1430, users 1435, and/or venders 1440, for example.
  • the inventory module 1420 may generate orders using an ordering engine 1445.
  • FIG. 15 depicts an exemplary GUI screenshot during a three-dimensional facial acquisition.
  • an alignment cross-hair 1505 is superimposed upon a user's real time image 1510.
  • Instructions 1515 are presented to the user on the left-hand side of the screen.
  • the user may be instructed to center the user's face upon the cross-hair and to push a "confirm" button 1520 when ready.
  • the exemplary virtual fitting station may then perform a 3D scan of the user's face 1510.
  • the user may be instructed to change positions for an additional 3D facial scan. For example, the user may be asked to open the user's mouth.
  • the system may use this information to determine the quality of fit, when the user may be talking, for example.
  • a user's actual body movement may be used to generate a dynamic facial model.
  • Exemplary dynamic modeling systems are described, for example, with reference to at least FIG. 3 in U.S. patent application, serial number 13/839,056, entitled “System and Method for Selecting a Respirator,” filed on March 15, 2013, the entire disclosure of which is incorporated herein by reference.
  • FIG. 16 depicts an exemplary depiction of the facial model compared with a user's face.
  • an exemplary GUI screenshot 1600 is depicted comparing the computer facial model 1605 of two users to an exemplary virtual fitting system.
  • the screenshot 1600 depicts a capture image 1610 of each user's face and depicts the captured image 1610 alongside the facial model 1605.
  • the system may also compute an error assessment image 1615.
  • the error assessment image 1615 may show the uncertainty of the model as a function of facial position.
  • the virtual fitting system may permit the user to reject the model and return to a 3D facial scanning stage depicted in FIGS. 15-15.
  • a new 3D facial model may be acquired by a 3D facial scanner, for example.
  • the user may then be queried as to the type of PPE device the user wishes to fit.
  • the virtual fitting system may then retrieve, from a previously created database, all PPE devices that match the user's criteria.
  • the virtual fitting station may proceed to computationally determine fitting metrics for each of the matched PPE devices.
  • the station may then sort the matched PPE devices according the computed fitting metrics. For example, the virtual fitting station may present to the user, a screen display depicting the matched PPE device having the best fitting metrics.
  • FIG. 17 depicts an exemplary GUI screenshot of a PPE device presentation to a user.
  • a virtual fitting device may first ask the user to select from among the matched fitting devices before computing fitting metrics on that device.
  • a virtual fitting system may present a screenshot containing all the matched devices along with the computed fitting metrics associated with each device.
  • the screenshot may present all matching devices sorted in decreasing order of a computed fitting metric.
  • FIG. 18 depicts an exemplary GUI screenshot of the presentation of a fitting metric resulting from the computation of a virtual fit.
  • This figure depicts an exemplary GUI screenshot 1800 displaying a representation of the facial model 1805 wearing a selected PPE device 1810.
  • the screenshot 1800 also depicts a graphic 1815 demonstration one of the computed fitting metrics.
  • the fitting metric displayed is the distance of the PPE device from the face, when worn.
  • the metric displayed may be user selected via a radio button, for example.
  • the metric displayed may be user selectable via a drop-down menu.
  • the user may select a different PPE device from a drop-down menu 1820 on this display screen 1800.
  • the user may change the vantage point of the display metric. For example, the user may want to see the fitting metric display from a close-up perspective. Or perhaps the user may want to see the chin fit of the PPE device, and therefore may rotate the display so as to better see the chin.
  • the user may be able to control the coloration of the displayed metric. For example, the use may change the distance range of coloration from zero inches to 3 inches. A user may want to color only those regions that the PPE device fits between one-quarter of an inch to one-half of an inch, for example.
  • FIG. 19 depicts an exemplary GUI screenshot of the presentation of a facial sealing location of a PPE device as computed during a virtual fit.
  • the exemplary FIG. 19 screenshot 600 depicts another exemplary fitting metric.
  • the contact line 605 between the PPE device and the user's face is shown.
  • the virtual fitting station may depict a graphic depicting a predicted comfort level for each contacted location of a user's face, for example.
  • a virtual fitting station may compute the expected level of heat a user might expect to experience wearing a particular PPE device.
  • a virtual fitting station may compute a seal quality for a virtually- fit PPE device.
  • a virtual fitting station may present a graphic in which the contact line is color-coded indicating the quality of fit at the various contact locations.
  • Exemplary methods of computing a PPE-facial seal metric are described, for example, with reference to at least FIG. 5 in US patent application 61/814,905 titled “System and Method for Selecting PPE Fit,” filed on April 23, 2013, the entire disclosure of which is herein incorporated by reference.
  • FIG. 20 depicts an exemplary GUI screenshot of a sealing metric of a PPE device as computed during a virtual fit.
  • an exemplary GUI screenshot 2000 shows a user virtually wearing a PPE device 2005.
  • the screenshot 2000 also depicts a computed fitting metric, in this example, the quality of a PPE-facial seal 2010.
  • the quality of the PPE-facial seal 2000 may be color coded, as depicted here.
  • the quality of the PPE- facial seal may be indicated by a single numerical metric 2012, for example.
  • the screenshot may permit the user to select from a list of PPE devices via a graphical selection area 715. These PPE devices may be have been selected by the virtual fitting station based upon matching criteria.
  • the matching criteria may have been supplied by the user.
  • the matching criteria may have been predetermined by the user's employer, for example.
  • the user may be able to select a different size of the PPE device by selecting among a series of size buttons 2025.
  • the user may be able to navigate forward and backwards through the various user screens using navigation buttons, for example.
  • FIG. 21 depicts an exemplary GUI screenshot of a dead-space metric of a
  • a screenshot 2100 includes an image of a user wearing a PPE device 2105, and a graphic 2110 displaying a computed fitting metric.
  • the displayed fitting metric is a dead-space score 2115.
  • the dead-space score may be displayed both graphically 2115 and using a single numeric metric 2120 as depicted here.
  • only the graphical display may be presented.
  • only a numeric metric will be presented.
  • more than one metric may simultaneously be presented to the user via a display screen.
  • the dead-space score may be low.
  • the dead-space score may also be low.
  • too much dead-space may not be considered problematic, for example.
  • too much dead-space may facilitate mask collapse during inhalation, for example.
  • Exemplary dead-space metric methods are described, for example, with reference to at least FIG. 8 in US patent application, serial number 13/839,186, entitled “System and Method for Selecting a Respirator,” filed on March 15, 2013, the entire disclosure of which is incorporated herein by reference.
  • FIG. 22 depicts a block diagram of an exemplary virtual fitting system.
  • a block diagram 2200 of an exemplary virtual PPE device fitting includes a face capture module 2205.
  • the face capture module may be used to capture users' 3D topological information.
  • the topological information may then be imported a mask fitting system 2210.
  • the mask fitting system may have a processor that may generate a 3D facial model and store such 3D facial models in a face model database 2215.
  • An employer may keep facial models of its employees in such a database, for example, so that employees' faces need not be reacquired.
  • a PPE product database 2220 may also be maintained by the exemplary virtual fitting station 2200.
  • the mask fitting system 2210 may retrieve both a facial model from the face model database 2215 and a PPE device model from the PPE product database 2220, for example. The mask fitting system may then use a fitting algorithm 2225 to compute fitting metrics for the facial model-PPE configuration. A fitting score 2230 may be output for a user's review. The mask fitting system 2210 may also compute a recommendation 2225 for the employee.
  • FIG. 23 depicts a flow chart of an exemplary virtual fitting system.
  • FIG. 23 depicts an exemplary virtual mask- fitting method 2300 from the vantage point of a processor in the mask fitting system 2210.
  • the processor begins by presenting to an alignment image for the user to align his live image to an alignment mark 2305.
  • the processor then waits until the user hits a "proceed" button 2310. If the user does not push the proceed button the processor continues to present the alignment screen 2305. If the user does push the
  • the processor then instructs the 3D scan tool to acquire the user's facial topological information 2315.
  • the processor then creates a 3D facial model using the acquired 3D facial information 2320.
  • the computer evaluates whether the model is acceptable 2325. If the model is not acceptable, the processor returns to step 2305 and presents the user an alignment screen. If the processor determines that the model is acceptable, however, the processor then requests the PPE criteria from the user 2330. The processor then waits for the user to input the PPE criteria 2335. When the processor receives the user input PPE criteria, the processor retrieves the first matching PPE device model 2340.
  • the processor computes fitting metrics during this step.
  • the computer assesses whether any more matching PPE models have yet to be virtually fit 2350. If more matching PPE device models have yet to be virtually fit, the processor returns to step 1040 and retrieves the next matching PPE device model. If all of the matching PPE devices have already been virtually fit, however, the processor then presents the PPE devices to the user 2355.
  • the PPE device devices may be presented in an order of decreasing computed fit metric, for example.
  • the processor then waits for a user to select a PPE device from the presented list 2360. When the user selects a PPE device, the processor then presents the user with a display of the computed data for the selected PPE device along with an image of the user wearing the selected PPE device2365.
  • the processor then asks the user if the user is satisfied with the selected PPE device 2370. If the user is not satisfied with the selected PPE device, the processor returns to step 2355 and again presents to the user the sorted list of all matching devices. If, however, the user is satisfied with the selected device, the processor finishes the exemplary method.
  • an exemplary virtually PPE fitting system may be performed on the cloud as a System as a Service (SaaS).
  • SaaS System as a Service
  • a virtual fitting system GUI may run remotely over the internet, for example.
  • the GUI may operate locally, while the processor computes the fitting metrics remotely.
  • the virtual PPE fitting system may fit a human body part, such as for example a hand.
  • Some embodiments may project the users actual real-time face upon the computed facial model. For example, the user may be moving in real time, and the selected PPE device may be superimposed upon the user's face creating a real-time virtual fit.
  • the Appendix details many various additional GUI aspects of an exemplary virtual fitting system.
  • Various embodiments may use more or less menton area elevation data from both the virtual masks and the virtual faces to optimally fit the menton region.
  • a fit-quality metric may be determined using the sealing surface periphery data. For example, the sealing surface data that may circumscribe the mouth and nose may be used to determine a fit-quality metric.
  • a dead-space volume that volume between the virtual facial elevations and the virtual mask interior elevations may be used in determining a quality metric.
  • a computer program product tangibly embodied in a computer readable medium and containing instructions that, when executed, cause a processor to perform operations to determine the fit of a virtual mask to a virtual face, the operations may include receiving facial elevation data corresponding to a face of a person.
  • operations may include retrieving from data memory locations mask elevation data corresponding to a facial mating surface of a mask model.
  • operations may include fitting a chin position of the retrieved mask elevation data to a chin region of the received facial elevation data.
  • operations may include rotating the retrieved mask elevation data toward or away from the received facial elevation data to a nose-bridge region of the received facial elevation data, while maintaining the fitted chin position. In some embodiments, operations may include calculating a quality of a fit between the retrieved mask elevation data and the received facial elevation data at the rotated position.
  • PPE Personal Protection Equipment
  • respirators are widely used by persons working in extreme environments.
  • a hazardous atmosphere may be one with excessive dust or particulate contamination, for example.
  • a hazardous atmosphere may be one with chemical vapors present.
  • Some nuclear facilities may have radioactive gas in the working atmosphere. Hospitals may present biological contaminants in the atmosphere.
  • respirators may be used by the workers who are exposed to such atmospheres.
  • Some employers may require certain employees to wear personal protection equipment to protect the employees from exposure to the hazards.
  • the quality of fit of personal protection equipment affects the level of protection that may be provided by the equipment. For example, if a respirator fits improperly, the employee wearing the respirator may be exposed to the hazard. If the quality of fit is poor, a respirator may be uncomfortable to wear. Many different types of respirators or other PPE devices are available in the marketplace. And each different piece of equipment may have a different presentation to the wearer. It can be a time consuming job to evaluate the many different equipment devices for the purpose of finding one with a good seal, and yet still is comfortable to wear.
  • Apparatus and associated methods may relate to a system for predicting a respirator fit by comparing a respirator model in a deformed state to a specific facial model.
  • an internal measurement may be calculated between an inside part of the respirator model and the facial model. The internal measurement may be compared against a predetermined threshold to determine a fit of the respirator model, for example.
  • the internal measurement may be a distance and/or a volume between the respirator and facial model.
  • a 3D representation of the respirator model may be displayed upon a 3D representation of the facial model.
  • a color-coded facial display may characterize areas of comfort and discomfort with respect to the respirator model.
  • an image capture device may generate point cloud data of a body part model and a PPE model.
  • an image capture device may generate point cloud data of a facial area and a respirator.
  • point cloud data may be used to overlay a respirator model on a facial model to determine a virtual placement of the respirator model on the facial model.
  • a rigid registration method may be used to align point clouds of the facial model and the respirator model.
  • identifying feature points of the body part model e.g., nose, mouth
  • a contact line may be determined upon the facial area.
  • Determination of the contact line may provide identification of a portion of a facial area that aligns with the inside part of the respirator.
  • Various embodiments may achieve one or more advantages. For example, some embodiments may objectively determine whether the hidden, inside part of the respirator will make contact with a portion of the facial area.
  • a predetermined deformation parameter of an outside part (e.g., outside surface) of the respirator may be attributed to a corresponding inside part (e.g., inside surface) of the respirator.
  • the deformation parameter may be determined at each of the vertices or point cloud of the outside and inside parts of the respirator.
  • internal measurements may be made between each deformed point cloud or each of the vertices and an aligned part of the facial area to determine a respirator fit.
  • Apparatus and associated methods may relate to a system for predicting a respirator fit by comparing a specific respirator model to a specific facial model in a dynamic position.
  • one or more dynamic positions may be generated by actual user movement and/or simulated user movement.
  • a facial model may be generated by altering a static model in view of actual and/or simulated movements.
  • a facial model may be compared against a variety of respirator models from a respirator model database.
  • a 3D representation of the respirator model may be displayed upon a 3D representation of the facial model.
  • Some embodiments may predict a realistic fit of a respirator to a facial area by modeling the facial area in one or more dynamic positions.
  • the dynamic positions may be characteristic of facial movements that a user may undergo while wearing the respective PPE, such as for example an open mouth, a raising head, or a bowing head.
  • the dynamic positions may be extreme facial movements.
  • FIG. 24 depicts an overview of an exemplary respirator selection system.
  • a system 2400 for selecting a PPE is illustrated, where the system 2400 may provide recommendations to a user on which PPE will provide an optimal fit, or provide a best fit among available PPE.
  • a fit level of the PPE upon the user may be determined by a variety of factors determined by the user, employee, and/or manufacturer. For example, an amount of predicted leakage of ambient air through a sealing edge of the respirator may be a determined factor for determining a fit level of the respirator on a facial area of the user.
  • the respective respirator may be given a low score and/or a non recommendation.
  • a force applied by the respirator upon a facial area of the user may be a determinant factor on a recommendation of a particular respirator type of size. If the respirator is predicted to apply pressure to the facial area at a rate or force exceeding a threshold, the respirator may be given a low score and/or a non recommendation because of a possible low comfort level provided to the user by the respirator, for example.
  • the system 2400 may provide fit recommendations or scores based upon captured and analyzed images of the user body part (e.g., facial area) and PPE (e.g., respirator).
  • the system 2400 include one or more image capture devices 2405 for capturing representations of a user body part 2410 and/or a type of PPE 2415.
  • the user body part 2410 is a user facial area.
  • the PPE 2415 may be a respirator, for example.
  • a series of two-dimensional (2D) images may be captured by the image capture device 2405 from which a three- dimensional (3D) image may be assembled.
  • a 2D image may be used to determine a PPE fit.
  • the image capture device may capture a 3D representation of the body part 2410 and/or the PPE 2415.
  • facial coordinate data representative of a shape of a facial area and respirator coordinate data representative of a respirator shape may be analyzed to provide a fit recommendation.
  • the system 2400 may load previously captured and/or generated body parts 2410 and/or PPE 2415.
  • the system 2400 may be used for selecting a variety of PPE 2415 to be worn on the intended body part 2410.
  • the system 2400 may predictively choose an optimal fitting glove to fit a user hand.
  • the system may choose an optimally fitting helmet to fit a head of a user.
  • several respirator point cloud data sets each indicative of a specific size and shape respirator 2415 may be stored in a database 2420. For example, each respirator that an employer offers to employees may be analyzed with associated
  • the representative point cloud data may be stored in a database 2420.
  • the point cloud data may include x, y, z coordinates which may be assembled to form a 3D image of the intended PPE 2415 and/or user body part 2410.
  • the database 2420 may be accessible over a wide-area network (e.g., Internet) to permit a wide selection of PPE 2415 to users without the need to personally capture data representative of each eligible PPE 2415.
  • a wide-area network e.g., Internet
  • a comparator module 2425 compares the PPE 2415 with the body part 2410 to determine whether the PPE 2415 will properly fit the respective body part 2410.
  • the PPE 2415 is overlaid upon the body part 2410.
  • a point cloud and/or vertices may be aligned between the PPE 2415 and the body part 2410.
  • the comparator module 2425 uses a set of predetermined rules from a rules library 2430 to determine whether the PPE 2415 properly fits the body part 2410.
  • the rules may require the sealing edge of a respirator not to be in contact with the mouth of the user.
  • the rules may require the respirator to have a surface area as large as the respirator-receiving portion of the facial area of the user.
  • the rules may identify a captured body part, such as for example a facial area, and direct the comparator module to only compare respirators from the database with the body part.
  • the rules may identify a captured body part, such as for example a hand, and direct the comparator module to only compare gloves from the database with the body part and not to compare respirators with the captured body part (e.g., hand).
  • a simulator module 2435 may display the fit.
  • the simulator module 2435 may display a representation of the respirator worn by the specific facial area of the user.
  • a predicted tightness or looseness of the PPE 2415 relative the body part 2410 may be emphasized in the simulator module 2435.
  • a predicted leakage between the sealing edge of the respirator and the facial area may be emphasized.
  • a report 2440 may be outputted to the user to assist in providing a recommendation on fit levels of each compared PPE 2415.
  • a list of evaluated PPE 2415 may be included in the report 2440 with each of the evaluated PPE 2415 having a score or fit level assigned.
  • only recommended PPE 2415 may be provided in the report 2440.
  • only the highest scoring three or five PPE 2415 may be provided in the report 2440.
  • FIG. 25 depicts a flowchart of an exemplary PPE selection system.
  • a determination may be made of whether one or more PPE models fit a specific body part.
  • the determination may be made by software operated on a processor module.
  • the software may determine which respirator of a plurality of respirator models from a respirator type database fits a specific facial area of a user and output a recommendation to a user.
  • data representing an exemplary body part may be captured as in step 2505.
  • the data may be captured by an image capture device.
  • an image capture device may scan a body part to build a 3D representation of the body part.
  • data representing the body part may be retrieved or computationally loaded.
  • the body part data may be retrieved from a body part database having specific body part shapes stored at an earlier date.
  • the data representative of a body part may be a generic body part computationally generated or morphed from one or more models.
  • the representative body part may be a facial area of a user.
  • data representing one or more types of PPE may be captured as in step 2510.
  • the data may be captured by an image capture device.
  • an image capture device may scan PPE to build a 3D representation of the PPE.
  • data representing the PPE may be retrieved or computationally loaded.
  • the PPE data may be retrieved from a PPE database having specific PPE shapes stored at an earlier date.
  • the representative PPE may be a respirator.
  • a particular type of PPE e.g., helmets, gloves, PPE
  • a PPE type including respirator models may be loaded if the intended body part may be a facial area.
  • a PPE type having gloves may be loaded if the intended body part may be a hand.
  • a first PPE model from the loaded relevant PPE type may be retrieved as in step 2520 to be compared via a comparator module with the captured body part as in step 2525.
  • Each PPE model may be distinguishable because of a size, shape, or other criteria which may affect the fit of the PPE on the user body part.
  • the comparison may determine whether the PPE model has a shape that will permit an acceptable fit over the shape of the body part.
  • the PPE model may be required to meet or exceed one or more thresholds or rules previously determined as indicative of proper or optimal fitting criteria.
  • the PPE model may be required to fit the body part in both static and dynamic states of the body part.
  • the PPE model may be simulated on the body part as in step 2535.
  • the PPE model and body part may be displayed to the user in a 3D
  • the user may rotate and pan a 3D representation of the simulated body part and PPE model.
  • the simulated representation may provide visual feedback to the user detailing areas of the PPE model that are not predicted or determined to fit the respective body part. For example, one or more colors may be overlaid upon the representation to indicate areas upon the body part that are predicted to be uncomfortable as a result of wearing the PPE model.
  • a blinking or flashing area may indicate an area of the PPE model that does not conform to a minimum threshold determined to be required to provide a proper and/or comfortable fit. For example, a portion of a sealing edge of a respirator may blink if a leak is predicted to be present in the respective portion.
  • the software may determine if there are any other PPE models in the chosen PPE type group that are to be evaluated against the respective body part as illustrated by step 2540. If so, the software cycles to a second PPE model as illustrated by step 2545 and repeats the process of steps 2525-2540. If there are no more PPE models from the PPE type group, a report and proposal may be generated for output to the user as illustrated in step 2550. In some exemplary embodiments, the report and proposal may include the top three PPE models that have the best fit with respect to the specific body part. In some exemplary embodiments, the top PPE models or all of the PPE models evaluated may be provided with a fit score to the user. In some exemplary embodiments, a different PPE type group may require comparison with the body part, in which case some or all of the process may be repeated.
  • Figure 26 depicts an overview of an exemplary PPE selection system.
  • the PPE selection system 2600 may be used to select an optimal fit PPE for a user body part during dynamic conditions of the user body part.
  • the PPE selection system 2600 includes a device type selection module 2605 for receiving a command from a user 2610.
  • the device type selection module 2605 sends commands to a PPE database 2615.
  • the PPE database 2615 may include a variety of types of PPE 2620, such as for example gloves, respirators, and helmets.
  • the device type selection module 2610 may relay a command 2610 indicative of a particular type of PPE 2620, such as for example a facial respirator.
  • the command 2610 may be indicative of a particular user body part 2625 to be matched with the PPE 2620 from the PPE database 2615.
  • the device type selection module 2605 may direct an image capture device (not shown) to capture a 2D or 3D image of the selected body part 2625.
  • the PPE 2620 may be modeled in a corresponding 3D shape.
  • one or more device range rules may define a capture range of the body part 2625 for the corresponding PPE 2620. For example, with half-mask respirators, the device range rules may define a capture range as the user face.
  • a capture range computing step may calculate a maximum facial area range that may accommodate the PPE and then correlate the range with each PPE to determine whether the respective PPE fits within the facial area range.
  • the user body part may be modeled using a static modeling module 2630.
  • the static modeling module generates a 3D model of the user body part to be used by a dynamic modeling module 2635.
  • the dynamic modeling module 2635 communicates with an input module 2640 for generating dynamic models of the user body part 2625.
  • the input module 2640 communicates actual body movement 2645 to the dynamic modeling module 2635 for generating a dynamic model using actual movement from the user.
  • the input module 2640 communicates simulated movement 2650 to the dynamic modeling module 2635 to generate a dynamic model based upon a simulated body part movement.
  • the dynamic model may then be compared with the PPE models 2620 from the PPE database 2615 by a comparator module 2655.
  • the comparator module 2655 may determine a fit level of the PPE model 2620 with the dynamic model from the dynamic modeling module 2635 based on a variety of predetermined criteria or rules. For example, the comparator module 2655 may evaluate a size of the PPE model 2620 with the dynamic model in an extreme position (e.g., open mouth) to determine whether the PPE (e.g., respirator) will fit the user body part (e.g., facial area) in the extreme position.
  • the calculated results of the comparator module 2655 may be summarized for output and visualization.
  • the comparator module 2655 may fit the PPE model candidate 2620 to the dynamic model set according to mapping rules. The comparator module 2655 may then calculate the difference between the PPE model candidate 2620 and dynamic model outputted from the dynamic modeling module 2635. According to a set of predetermined evaluation rules, a fit score of each PPE model 2620 may be provided relative the dynamic model. Lastly, the comparator module 2655 may output an optimal fit PPE 2620 based on simulated comfort and fit. In an exemplary embodiment, a respective fit of the PPE 2620 may be visualized by color coding for user.
  • the result from the comparator module 2655 may be outputted to a simulator module 2660 for display to a user through an output module.
  • the simulator module may graphically overlay the 3D PPE model 2620 upon a 3D representation of the user body part 2625 to illustrate to the user the PPE model 2620 being virtually worn on the user body part 2625.
  • a fit level, score, or color may accompany the graphical illustration for ease in interpreting the results.
  • the output module 2665 may comprise a display module. In some exemplary embodiments, the output module 2665 may comprise a printed report. In some exemplary embodiments, the report may provide 3D visual representations of the PPE virtually worn by the user. In some exemplary embodiments, the report may provide a detailed list of a fit level or score of each evaluated PPE with respect to a region of interest of the user. In some exemplary embodiments, the report may provide a color-coded graphical representation of a PPE virtual fit on the user. In some exemplary embodiments the color-coded graphical representation may illustrate, through color-coding, different levels of pressure as applied to the user by the PPE when virtually worn.
  • FIG. 27 depicts a flowchart of an exemplary process of a static modeling module.
  • a static modeling module 2700 may be used to generate an objective static 3D model of a body part.
  • the static model may be used in dynamic modeling processes.
  • the PPE may be compared directly to the static model if dynamic comparison is not necessary.
  • a static model of a facial area may be generated with the static modeling module 2700.
  • the generated static model may be in 2D form.
  • the module 2700 first captures a region of interest (ROI) point cloud of a user as in step 2705.
  • the ROI may be the portion of the body that corresponds to the evaluated PPE.
  • the ROI may be a facial area of the user.
  • the point cloud may include x, y, z coordinates assembled to form a 3D image of the respective body part.
  • a generic model may also generated as illustrated in step 2710 to generically match the body portion captured by the point cloud as in step 2705.
  • the generic model may be representative of a generic user face.
  • the generic model may be retrieved from a database of generic models.
  • a preliminary screening process may be completed to find a generic model being close in shape to the captured ROI.
  • Predetermined semantic information is attributed to the generic model as in step 2715.
  • the semantic information may be distinguishable body feature points of the corresponding body part.
  • a facial area may include semantic information associated with the eyes, ears, mouth corners, and a nose tip.
  • the semantic information may be attributed to the vertices of the generic model, for example.
  • a set of rules which define the semantic information may include MPEG4 format definition rules.
  • the point cloud of the ROI is then overlaid on the generic face model by a rigid method as in step 2720.
  • the rigid method may include a rigid registration or alignment of vertices of the ROI and vertices of the generic model.
  • the aligned vertices may correspond to proximally similar or equivalent locations on the modeled body part.
  • the nose portion of the point cloud of the ROI may be aligned with nose portion of the generic model.
  • the module 2700 determines whether the vertices of the point cloud align or match to an acceptable level or threshold as illustrated in step 2725.
  • a non-rigid registration method may include blending the non-aligning vertices of the generic model with neighboring vertices.
  • certain vertices of the generic model may be moved a predetermined allowable distance to reach an alignment with the point cloud of the ROI.
  • the semantic information of each vertex on the generic face model may be attributed to the point cloud.
  • the vertices of the point cloud may receive the semantic information and be stored within the properties of the point cloud such that each of the points in the point cloud may include identification properties corresponding to a location of the point in the point cloud.
  • a point of the point cloud located at a position corresponding to a nose tip may include semantic information identifying the point as "nose tip”.
  • the static model having the point cloud with semantic information may then be outputted as in step 2740.
  • the static model may be outputted to the dynamic modeling module.
  • the static model may be outputted to a comparator module.
  • the static model may be outputted to a simulator module.
  • the static model may be a 3D representation.
  • FIG. 28A depicts a flowchart of an exemplary dynamic modeling module using actual body movement.
  • a dynamic modeling module 2800 uses actual body movement to generate a dynamic model as described with reference to Figure 26.
  • the dynamic model may be generated in 3D form.
  • a user performs actual body movement as in step 2805 and an image capture device captures the body movement as in step 2810.
  • the body movement performed may correlate with body movement commonly performed while wearing associated PPE. For example, the user may speak a variety of phrases when fitting a user with a respirator since it may be common for a user to speak or move their mouth while wearing the respirator.
  • a key frame may be extracted from the captured sequence as in step 2815.
  • the key frame may be an image reflecting a particular user movement, for example.
  • the key frame may simply be a generic reference or starting image.
  • the user may then manually mark feature points on the selected, first key frame as in step 2820.
  • the feature points may correspond with distinguishable features on the body part captured. For example, a nose or mouth may be feature points for a captured facial area.
  • the user manually selects the feature points by visually selecting the feature points on a computer display.
  • the user manually selects the feature points by selecting body coordinates predetermined to be associated with the respective feature point.
  • the selection of the feature points may be automated via an image recognition software or device.
  • the feature points may be appointed identifying information, such as for example semantic information.
  • a tracking method may be performed to identify and mark feature points on all key frames based on the selected feature points of the first key frame as in step 2825.
  • the tracking method may track the feature points via proximity of similar vertices in neighboring key frames.
  • the tracking method may be automatically performed by the dynamic module 2800.
  • One of the key frames having feature points may then be selected and the feature points corresponded to a static 3D model as in step 2830.
  • the static 3D model may be generated according to the detailed process exemplified in Figure 27.
  • the feature points may be linked to similarly located points from the point cloud of the static model such that properties of the points from the point cloud of the static model may be transferred to the feature points of the key frame, for example.
  • a morphed model may then be generated by morphing the static model to a facial position of the key frame.
  • the point cloud of the static model may be altered to a proximal location of the feature points. If the key frame illustrates a user having an open mouth, the static model and associated point cloud may be altered to reflect an open mouth morphed static model.
  • the morphable model may be generated by performing rigid and/or non-rigid registration methods with vertices or points between the key frame and the static model.
  • the module 2800 determines whether there are additional key frames to analyze. If there are more key frames to analyze, then the module cycles to the next key frame as in step 2845 and returns to step 2830. If there are no more key frames to analyze, then a dynamic model may be outputted as in step 2850. In an exemplary embodiment, the dynamic model may be outputted to a comparator module for comparing the PPE model with the dynamic model to determine whether the PPE model fits the dynamic model. In an exemplary embodiment, the dynamic model may be outputted as a 3D model set for all captured key frames.
  • Figure 28B depicts an exemplary graphical view of the dynamic modeling module of Figure 28A.
  • a key frame 2855 having feature points manually marked on the key frame as described with reference to step 2820 of Figure 28A.
  • the static model 2860 may be generated by the static modeling module as described with reference to Figure 27.
  • the point cloud of the static model may be located along similar facial features as the feature points of the key frame.
  • a morphable model 2865 may be generated by combining the key frame and the static model as described with reference to step 2835 of Figure 28A.
  • Figure 29A depicts a flowchart of an exemplary dynamic modeling module using simulated body movement.
  • a dynamic modeling module 2900 uses simulated body movement to generate a dynamic model as described with reference to Figure 26.
  • the dynamic model may be generated in 3D form.
  • an extreme movement set is defined as in step 2905.
  • the extreme movement set may be defined by a user in an exemplary embodiment. In other exemplary
  • the extreme movement set may be defined by the PPE manufacturer, regulatory agency, and/or employer.
  • extreme movements may include raising the head, bowing the head, speaking, and/or opening the mouth.
  • the extreme movement set may be defined according to actions a user may typically undergo while wearing the respective PPE.
  • a first extreme movement may be selected from the set as in step 2910.
  • Feature points affected by the selected extreme movement are marked or identified on a static model. For example, if the extreme movement selected mimics an open mouth, feature points surrounding a mouth of the static model may be marked or identified. In an exemplary embodiment, the feature points are linked to corresponding proximal vertices.
  • the static model may be generated by a process as exemplified with reference to Figure 27, for example.
  • An influence zone of each feature point may also be defined on the static model as illustrated by step 2920.
  • the influence zone may be a proximal area of each feature point that may be affected by movement of the respective vertex.
  • the feature points and/or feature point influence area may correspond to predetermined data points of an MPEG 4 standard.
  • the feature points may include semantic information.
  • the user manually selects the feature points by selecting body coordinates predetermined to be associated with the respective feature point.
  • the selection of the feature points may be automated via an image recognition software or device.
  • the feature points may be appointed identifying information, such as for example semantic information.
  • the maximum feature point position is also defined as in step 2925.
  • the maximum feature point may correspond to a maximum distance and x, y, z coordinate location of the feature point away from a normal or current location of the feature point on the static model.
  • the vertices and linked feature points are then displaced to the maximum position as defined by the extreme movement as in step 2930 and a morphed model is formed as in step 2935.
  • a prior defined deformation function affect, neighbor related points on static model are displaced to a new position.
  • the displacement position of neighbor points can be calculated by:
  • D vertex may be the displacement position of neighbors
  • Dpp displacement of feature points
  • H is the deformation function.
  • the influence zone of each feature point may also be blended or altered according to linked feature point movement.
  • neighboring vertices may be blended by a weighted sum.
  • a deformation function may be defined as:
  • T is the radius of the area that applies the deformation
  • ⁇ with the scope of (0, 1) is a parameter to adjust the deformation degree; if ⁇ is close to 1, the deformation will be smooth, and if ⁇ is close to 0, the deformation will be sharp.
  • step 2940 the module 2900 determines whether there are additional key frames to analyze. If there are more extreme movements to analyze, the module cycles to the next extreme movement as in step 2945 and returns to step 2910. If there are no more extreme movements to analyze, a dynamic model may be outputted as in step 2950. In an exemplary embodiment, the dynamic model may be outputted to a comparator module for comparing the PPE model with the dynamic model to determine whether the PPE model fits the dynamic model. In an exemplary embodiment, the dynamic model may be outputted as a 3D model set for all captured key frames.
  • Figure 29B depicts a graphical view of the exemplary dynamic modeling module of Figure 29A.
  • the exemplary process includes a first defined extreme movement 2955, such as for example an open mouth.
  • a static model 2960 may be imported, such as for example the static model generated by the static model module with reference to Figure 27.
  • the feature points 2965 and influence zones 2970 corresponding to the extreme movement 2955 are marked on the static model 2960.
  • the feature points may be located by correspondence with an MPEG 4 standard.
  • a deformation function 2975 may be executed to morph the static model by displacing the feature points and influence zone according to the maximum position as defined by the extreme movement.
  • a morphed dynamic model 2980 is outputted.
  • the dynamic model may be generated in 3D form.
  • the dynamic model has a body position correlating to the defined extreme movement.
  • Figure 30 depicts an overview of another exemplary PPE selection system.
  • a PPE selection system 3000 may be used to select an optimal fit PPE for a user body part based on an internal space measured between the PPE and the body part.
  • the internal space may be measured during a deformed state of the PPE.
  • the PPE selection system 3000 includes a device type selection module 3005 for receiving a command from a user 3010.
  • the device type selection module 3005 sends commands to a PPE database 3015.
  • the PPE database 3015 may include a variety of types of PPE 3020, such as for example gloves, respirators, and helmets.
  • the device type selection module 3010 may relay a command 3010 indicative of a particular type of PPE 3020, such as for example a facial respirator.
  • the command 3010 may be indicative of a particular user body part 3025 to be matched with the PPE 3020 from the PPE database 3015.
  • the device type selection module 3005 may direct an image capture device (not shown) to capture a 2D or 3D image of the selected body part 3025.
  • the PPE 3020 may be modeled in a corresponding 3D shape.
  • one or more device range rules may define a capture range of the body part 3025 for the corresponding PPE 3020. For example, with half-mask respirators, the device range rules may define a capture range as the user face.
  • a capture range computing step may calculate a maximum facial area range that may accommodate the PPE 3020 and then correlate the range with each PPE 3020 to determine whether the respective PPE 3020 fits within the facial area range.
  • the user body part 3025 may be modeled using a contact line module 3030.
  • the contact line module 3030 determines a contact line of the edge of the PPE 3020 on the body part 3025 of the user.
  • a respirator sealing edge may be defined as the contact line since the sealing edge may be the primary portion of the respirator that makes contact with the user facial area.
  • the contact line may be determined by capturing an image of the user wearing the PPE 3020 and not wearing the PPE 3020, and then using a subtractive function to find the contact line.
  • the contact line may be found by capturing a 2D or 3D image of the user wearing and not wearing the PPE 3020.
  • the contact line may be determined using previously captured models of users and/or PPE 3020.
  • the previously captured models may be aligned using a rigid or non-rigid registration method to calculate a contact line.
  • the portion of the body part 3025 confined by the contact line may be determined.
  • a portion of a face confined and within the contact line of a respirator may include a portion of a nose and a mouth.
  • a deformation module 3035 may then be used to deform the PPE 3020.
  • the PPE 3020 may be deformed according to a set of predetermined rules. For example, if a respirator is known to partially collapse inwards a certain percentage during wear, the PPE 3020 model may be deformed an amount or distance equivalent to a calculated standard collapse of an in-use respirator.
  • the degree of deformation may be determined by a maximum flex permissible by the construction of the PPE 3020.
  • a deformation of an inside surface or part of the PPE 3020 may be determined or calculated from a deformation of an outside surface or part of the PPE 3020.
  • a deformation of an outside part of the PPE 3020 may be computed by comparing the outside part of the PPE 3020 to a deformation of the inside part of the PPE 3020.
  • a comparator module 3055 may determine a fit level of the deformed PPE 3020 model 3020 with respect to the portion of the body part 3025 internal or confined by the contact line.
  • an internal measurement may be made between the internal surface of the PPE 3020 and the portion of the body part 3025 confined or internal to the contact line. For example, a distance between an inside surface of a respirator and a portion of a user face perpendicular to the inside surface may be calculated while the respirator is in the deformed state.
  • the internal measurement may be a distance between the PPE 3020 and the body part 3025.
  • the internal measurement may be an internal volume confined between the inside of the PPE 3020 and the corresponding body part 3025.
  • the internal measurement may be compared against a predetermined threshold to determine whether the PPE 3020 meets predetermined fit criteria. For example, if the predetermined threshold is not large enough, the PPE 3020 may be disqualified from an acceptable fit category of PPE 3020.
  • the calculated results of the comparator module 3040 may be summarized for output and visualization.
  • the internal measurement may use an implicit function to calculate a distance between the inside part of the PPE 3020 and the
  • a Gaussian smooth function may then be applied to the distance calculation, for example.
  • a color-coded result of the internal measurement may be outputted to a user.
  • the comparator module 3040 may fit the PPE 3020 model candidate 3020 to the user body part 3025 set according to mapping rules. According to a set of predetermined evaluation rules, a fit score of each PPE 3020 model 3020 may be provided relative the user body part 3025. Lastly, the comparator module 3040 may output an optimal fit PPE 3020 based on simulated comfort and fit. In an exemplary embodiment, a respective fit of the PPE 3020 may be visualized by color coding for user. [00137] In an exemplary embodiment, the result from the comparator module 3040 may be outputted to a simulator module 3045 for display to a user through an output module.
  • the simulator module may graphically overlay the 3D PPE 3020 model 3020 upon a 3D representation of the user body part 3025 to illustrate to the user the PPE 3020 model 3020 being virtually worn on the user body part 3025.
  • a fit level, score, or color may accompany the graphical illustration for ease in interpreting the results.
  • the output module 3050 may comprise a display module.
  • the output module 3050 may comprise a printed report.
  • the report may provide 3D visual representations of the PPE 3020 device virtually worn by the user.
  • the report may provide a detailed list of a fit level or score of each evaluated PPE 3020 device with respect to a region of interest of the user.
  • the report may provide a color-coded graphical representation of a PPE 3020 device virtual fit on the user.
  • the color-coded graphical representation may illustrate, through color-coding, different levels of pressure as applied to the user by the PPE 3020 device when virtually worn.
  • Figure 31 depicts a flowchart of another exemplary PPE selection system.
  • an optimal fit PPE for a user body part based on an internal space measured between the PPE and the body part.
  • the system provides a method of measuring an internal or hidden space between a PPE and an associated body part to generate an objective fit of the PPE on the body part.
  • a point cloud of an outside part of a PPE may be captured during a deformed and non deformed state to determine a position or shape of an inside part of a PPE.
  • a point cloud may be captured of a region of interest (ROI) of a body part of the user as in step
  • ROI region of interest
  • a point cloud may be captured of a facial area of the user.
  • a point cloud may be captured of a PPE as in step 3110, such as for example a respirator.
  • a point cloud may be captured both of the outside part (e.g., outside surface) and the inside part (e.g., inside surface) of the PPE.
  • a point cloud may also be captured of the ROI with the PPE being worn as in step 3115.
  • the point cloud data may include x, y, z coordinates which may be assembled to form a 3D image of the intended PPE and/or user body part.
  • a point cloud may include semantic information or other identifying feature points of the user body part and/or PPE.
  • an image capture device may directly capture a 2D or 3D image of the selected body part ROI and/or PPE.
  • previously captured 2D or 3D images of body part ROI and/or PPE may be used.
  • captured point cloud of people with and without device may be retrieved independently and compared to get placement information for the outside part of PPE. An estimate of placement and fit of the inside part of PPE may be made, for example.
  • a contact line may also be defined as in step 3120.
  • the contact line may be the point or edge that the PPE makes contact with the user ROI, such as for example a sealing edge of a respirator on a face of a user. Once the contact line is determined a portion of the ROI that is confined or within the contact line may be determined as will be described.
  • a deformation of the PPE may also be calculated, measured, or determined as in step 3125.
  • a deformation of an inside or outside part of the PPE may be calculated or measured based on a deformation of a respective outside or inside part of the PPE.
  • a degree of deformation may be predetermined by a manufacturer.
  • a degree of deformation may be determined by an employer based on common workplace practices.
  • the inside or outside part of the PPE may be used to generate the deformed PPE structure, thus only one of the inside or the outside part of the PPE may be needed.
  • the internal space between the PPE and the portion of the ROI confined by the contact line may then be measured as in step 3130.
  • the internal space may be determined based on a perpendicular distance between the PPE and the ROI.
  • the internal space may be determined by a contained volume between the PPE and the ROI.
  • the internal space may be measured while the PPE is in a deformed state.
  • a comparator module may determine whether a threshold has been met by the measured internal space as in step 3135. If a predetermined threshold has been met, then a positive recommendation may be outputted to a user as in step 3140. In an exemplary embodiment, a 3D visual representation of the PPE on the ROI may be displayed to the user. In another exemplary embodiment, the internal measurement may be displayed on the 3D visual representation. If a predetermined threshold has not been met, then a negative recommendation may be outputted to a user as in step 3145. For example, if the distance between an internal surface of a respirator and the beneath facial area does not meet a predetermined length, then the respirator may fail a fit test.
  • FIG 32A depicts a flowchart of another exemplary PPE selection system.
  • an optimal fit PPE for a user body part based on an internal space measured between the PPE and the body part.
  • the system 3200 may compute a fit of the PPE on the user body part based on previously captured point clouds of the PPE and/or body part.
  • point cloud data may be captured of the body part ROI (e.g., face) and the PPE (e.g., respirator) as in step 3205.
  • point cloud data may be captured of the ROI with the PPE being worn.
  • point cloud data of the PPE may be overlaid upon point cloud data of the ROI as in step 3210.
  • the point cloud data may be aligned using a rigid-registration method.
  • the rigid-registration method aligns feature points of the PPE and ROI.
  • the rigid-registration method aligns semantic information of the PPE and ROI.
  • the rigid- registration method aligns corresponding vertices of the PPE and ROI.
  • a contact line of the PPE on the ROI may be obtained as in step 3215.
  • the contact line may be the point or edge that the PPE makes contact with the user ROI, such as for example a sealing edge of a respirator on a face of a user.
  • the contact line may be visibly or
  • a center part of the contact line may be determined.
  • the center part of the contact line may be the center of a medial axis of the contact line.
  • the medial axis may be vertically oriented and separate left and right sides of the space confined by the contact line.
  • a center part of an inside part of the PPE may also be computationally determined and the center part of the PPE and the center part of the ROI are aligned.
  • rigid- registration methods may be used to obtain placement of the PPE on the ROI by aligning the center parts of the PPE and the ROI. Once the center parts of the ROI and PPE are aligned, corresponding points of the ROI and PPE may be determined and confirmed such as for making internal measurements.
  • a deformation of the PPE may be calculated, measured, or determined as in step 3230.
  • a deformation of an inside or outside part of the PPE may be calculated or measured based on a deformation of a respective outside or inside part of the PPE.
  • a degree of deformation may be predetermined by a manufacturer.
  • a degree of deformation may be determined by an employer based on common workplace practices.
  • the inside or outside part of the PPE may be used to generate the deformed PPE structure, thus only one of the inside or the outside part of the PPE may be needed.
  • the internal space between the PPE and the portion of the ROI confined by the contact line may then be measured as in step 3235.
  • the internal space may be determined based on a perpendicular distance between the PPE and the ROI.
  • the internal space may be determined by a contained volume between the PPE and the ROI.
  • the internal space may be measured while the PPE is in a deformed state.
  • a comparator module may determine whether a threshold has been met by the measured internal space as in step 3240. If a predetermined threshold has been met, then a positive recommendation may be outputted to a user as in step 3245. In an exemplary embodiment, a 3D visual representation of the PPE on the ROI may be displayed to the user. In another exemplary embodiment, the internal measurement may be displayed on the 3D visual representation. If a predetermined threshold has not been met, then a negative recommendation may be outputted to a user as in step 3250. For example, if the distance between an internal surface of a respirator and the beneath facial area does not meet a predetermined length, then the respirator may fail a fit test.
  • FIG 32B depicts an exemplary center part on a ROI as defined with reference to Figure 32A.
  • a region of interest ROI 3255 may be a body part that is to be protected, such as by a corresponding PPE.
  • the ROI 3255 may be a facial area of a user.
  • the ROI 3255 may be illustrated in a 3D form to a user.
  • the ROI 3255 includes point cloud data used in the construction of the 3D form and the fitting of the PPE.
  • a contact line 3260 may be defined on the ROI 3255, as previously defined with reference to step 3215 of Figure 32A.
  • the contact line 3260 may be peripheral edge of the PPE that makes contact with the ROI 3255, such as for example a sealing edge of a respirator.
  • the contact line 3260 may be computationally determined by comparing a user ROI while wearing and while not wearing a PPE.
  • the contact line 3260 may be manually drawn on the ROI by tracing a peripheral edge of the PPE worn on the ROI.
  • a center part 3265 of the contact line 3260 may also be defined, as previously defined with reference to step 3220 of Figure 32A.
  • the center part 3265 includes a medial axis 3270 and axis center 3275.
  • the medial axis 3270 may separate two-halves of the area of the ROI 3255 defined by the contact line 3260.
  • the medial axis 3270 may separate left and right halves of the area of the ROI 3255 defined by the contact line 3260.
  • the axis center 3275 may be the lengthwise center of the medial axis 3270.
  • a PPE center part including a PPE medial axis and PPE axis center are also defined on the PPE with reference to a contact edge (e.g., sealing edge of a respirator).
  • the PPE medial axis and PPE axis center of the PPE are then aligned with the medial axis 3270 and axis center 3275 of the ROI to determine a placement of the PPE on the ROI, as previously defined with reference to step 3225 of Figure 32A.
  • FIG. 33 depicts a flowchart of an exemplary deformation process.
  • a deformation module 3300 may determine a deformation of an inside part of PPE by correspondence with a deformation of an outside part of the PPE to obtain an internal measurement of the PPE and body part for determining whether the PPE fits the body part.
  • the module 3300 first captures point cloud data of the body part and PPE as in step 3305.
  • the point cloud data may be captured earlier in the process and relayed to the deformation module 3300.
  • the point cloud sets are overlaid upon each other as in step 3310 to fit the PPE to the body part.
  • the fitting process may be performed by other modules and the result relayed to the deformation module 3300.
  • the module 3300 may then obtain deformation parameters of the outside part of the PPE as in step 3315.
  • the outside part of the PPE may be an outside surface of the PPE with respect to the PPE being worn by a user.
  • the deformation parameters may be predetermined according to specific construction properties of the PPE.
  • the deformation parameters may be determined by functions, such as for example the deformation function described with reference to Figure 29A.
  • the PPE outside part may then be corresponded to the PPE inside part as in step 3320. For example, corresponding inside and outside part points may be correlated based upon a nearest distance between inside points and outside mesh nodes.
  • inside points or vertices determined to be physically affected by specific outside points or vertices are linked. For example, moving a point A on an outside part may correspondingly move a point B on an inside part of the PPE, and thus point A may be linked to some degree to point B.
  • the PPE inside part may be deformed according to the attributed deformation parameters linked to the respective inside part in step 3325.
  • the internal space between the inside part of the PPE and the portion of the ROI confined by the contact line may then be measured as in step 3335.
  • the internal space may be measured while the inside part of the PPE is in the deformed state.
  • the internal space may be determined based on a perpendicular distance between the PPE and the ROI.
  • the internal space may be determined by a contained volume between the PPE and the ROI.
  • a comparator module may determine whether a threshold has been met by the measured internal space as in step 3340. If a predetermined threshold has been met a positive recommendation may be outputted to a user as in step 3345. In an exemplary embodiment, a 3D visual representation of the PPE on the ROI may be displayed to the user. In another exemplary embodiment, the internal measurement may be displayed on the 3D visual representation. If a predetermined threshold has not been met, then a negative recommendation may be outputted to a user as in step 3350. For example, if the distance between an internal surface of a respirator and the beneath facial area does not meet a predetermined length, then the respirator may fail a fit test.
  • Figure 34 depicts a graphical representation of an exemplary color-coded display of a PPE fit.
  • a display 3400 may be outputted to a user by an output module for providing a visual recommendation of a PPE fit.
  • the display 3400 may be outputted on a computer screen.
  • the display 3400 may be outputted in a printable format.
  • the display 3400 includes a representation of the evaluated user body part 3405, for example a facial area.
  • the body part 3405 may be portrayed in 3D form.
  • the body part 3405 may be colored according to pressure distribution as applied on the body part 3405 by the PPE.
  • the PPE may be shown with the body part 3405.
  • the display 3400 includes a reference chart 3410 of the colors illustrated on the body part 3405 and values 3415 associated with each of the colors on the color chart 3410.
  • the values 3415 may represent ranges of pressure distribution, for example.
  • a user may visually determine whether a PPE would provide an acceptable fit by visualizing whether any areas upon the body part 3405 are a certain color. For example, if an area of the body part 3405 were colored red, a high degree of applied pressure may be applied to the body part 3405 by the respective PPE. For example, a respirator may fit tightly against a face of a user in a certain area. In an exemplary embodiment, if a certain color were displayed on the body part 3405 which would represent a threshold being exceeded, the respective PPE may be disqualified from further consideration with respect to the specific user.
  • shapes or symbols, rather than colors may be visually displayed on the body part 3405 to symbolize measured criteria.
  • a first shape may represent a first pressure applied to the body part 3405 by the PPE and a second color may represent a second pressure applied to the body part 3405 by the PPE.
  • a first color, shape, or pattern may be overlaid upon the body part 3405 to represent a first distance that the PPE is from the body part when virtually worn, and a second color, shape, or pattern may be overlaid upon the body part 3405 to represent a second distance that the PPE is from the body part when virtually worn.
  • FIG. 35 depicts a flowchart of an exemplary color-coded result generator.
  • a color-coded result generator 3500 may associated a set of fit results with one or more colors to provide a user with a quick method of determining whether a respective PPE would fit and/or be comfortable.
  • the fit results may be calculated and assigned to each of the points on the point cloud, such that each point in the point cloud of the ROI may have an associated fit result as illustrated in step 3505.
  • the result generator 3500 may calculate and assign the fit results or the fit results may be imported.
  • the fit results each include an applied pressure upon the body part by the PPE.
  • the fit results may include a pressure applied to a facial area by the respirator at each defined point or vertices.
  • the generator 3500 may correlate one or more colors to one or more predetermined ranges as in steps 3510 and 3515.
  • a first range of applied pressure values may be assigned a first color, for example a blue color.
  • a second range of applied pressure values may be assigned a second color, for example a green color.
  • the generator determines whether more colors are needed as in step 3520 and generates additional colors with assigned predetermined ranges as in step 3525.
  • a set of predetermined colors may be initially assigned that include all possible ranges, such as for example from -co to + ⁇ .
  • the color ranges may then be corresponded to the fit results as in step 3530.
  • a green color overlay on the human body part may represent an optimal match and a red color overlay on the human body part may represent a non optimal match.
  • the colors may represent how tight or loose PPE may be relative the human body part, such as for example red being shown for an area of the body part where the PPE product fits too tightly and green may be shown for an area of the body part where the PPE product fits too loosely.
  • the colors may then be overlaid on a body part ROI representation as in step 3535 and the result may be outputted to the user as in step 3540.
  • the body part ROI representation may be in 3D form.
  • An exemplary output is shown by display 3400 of Figure 34.
  • the system and method for automatically selecting a respirator may comprise predictive software that may capture a facial image and match the facial image to the closest form of a respirator model, type, and/or size.
  • the software may use a dynamic set of images and match the images to the flexibility of a respirator shape to predict an interface between the respirator model and the facial model. For example, the software may predict whether the interface between the respirator and the facial area will result in separation thus permitting leakage or breach in the sealing surface.
  • the image capture device may be a 3D digital scanner, such as for example one or more Kinect devices manufactured by Microsoft®.
  • the image capture device may be a still camera device.
  • the image capture device may be a video recorder device.
  • the image capture device may be a handheld unit.
  • the image capture device may be wirelessly connected to a processing module for receiving a scanned image from the image capture device and determining whether a scanned or modeled PPE fits a scanned or modeled body part.
  • the image capture device may be a low-cost item.
  • apparatus and methods may involve a digital image of a facial area of a user in a variety of facial positions.
  • a first facial position may be a grin or smile.
  • a second facial position may be the user voicing specific letters and/or sounds.
  • software may digitize the facial shape of the user in each of the facial positions to create a flexible electronic file, for example.
  • software may also store files having contours of respirators in both a static state and in a flexed state for comparison to facial shape files.
  • the software may match up a negative cavity of the respirator model with a positive face form of the facial area model to determine a fit level of a respirator or best fit respirator.
  • software may match the respirator and the facial area in both static and dynamic positions of the facial area and/or respirator to determine whether a respirator will fit in a variety of facial positions and/or flexed positions of the respirator.
  • an administrator may oversee a matching process of the respirator and a specific facial area.
  • an administrator at a workplace may oversee the matching process for each new employee.
  • each employee may undergo a matching process, such as for example via a pay per use web link.
  • a kiosk or vending machine may include software functionality to perform a matching process between one or more respirators and a specific facial shape.
  • a user may scan a user facial shape at a kiosk, and the kiosk may geometrically compare the facial shape of the user to a plurality of respirator models available for dispensing to find a respirator that most closely matches the facial shape of the user.
  • a population data gathering and storage system may be made available via scanning facial areas of users.
  • the facial shapes gathered and stored via the matching process may be used by respirator manufacturers to improve a respirator design such that newly manufactured respirators more closely match a common facial shape of persons commonly wearing the respirators.
  • the facial shapes gathered and stored via the matching process may be used by employers to provide insight on which respirators to stock in greater or less numbers.
  • a captured point cloud of a PPE and/or a user body part may be re-used in other PPE design.
  • a variety of body parts may be scanned and captured for being matched with respective clothing or garments.
  • a hand of a user may be scanned and stored as a data set such that a variety of glove models, types, and/or sizes may be compared against the hand of the user to find an optimal or best fit glove.
  • a head of a user may be scanned and stored as a data set such that a variety of helmet models, types, and/or sizes may be compared against the head of the user to find an optimal or best fit helmet.
  • a system and method for selecting a respirator may include a body modeling module for capturing an image(s) of a body part (e.g., facial area) of a user.
  • the image(s) may be used to generate a 3D model of the body part.
  • the system and method for selecting a respirator may include one or more product databases of PPE 3D models.
  • each product database may include PPE to be worn on a specific body part.
  • a respirator database may be associated with facial areas
  • a glove database may be associated with hands
  • a helmet database may be associated with heads.
  • the material properties of each specific PPE may also be stored with the specific PPE model.
  • the system and method for selecting a respirator may include a rule library illustrating a method of mapping 3D PPE models to a 3D human body part.
  • a rule library may include three types of rules, such as for example association rules, mapping rules, and evaluation rules.
  • association rules may define which related PPE 3D models from the product database are associated to a target body part.
  • respiratory products from product database may be associated to face models
  • footwear products from product databases may be associated to foot models.
  • mapping rules may define how the product model will be mounted to the body model, such as for example by mapping directions, forces, and/or deformations according to a material property.
  • evaluation rules may define how well the PPE fits the body part in accordance with a mapping result. For example, via dimensional comparison, a body dimension may be compared to a related product dimension range or pressure distribution during and after the product is mapped to the body part.
  • the system and method for selecting a respirator may include a 3D geometry matching module.
  • the matching module may calculate all differences between the 3D PPE models and the 3D human body model.
  • the geometry matching module may select a PPE part according to association rules, determine the difference with the mapping rules, summarize the difference according to the evaluation rules, and then propose a product model and/or size which may optimally fit a user.
  • a top three or top five best fitting products may be provided to the user.
  • the system and method for selecting a respirator may include a simulator module.
  • a simulator module may visualize to a user how well the PPE model fits on the body part model.
  • the simulator may display the human body part and PPE product in 3D representations.
  • color coding may be used to illustrate how well the PPE fits a human body part. For example, a green color overlay on the human body part model may represent an optimal match and a red color overlay on the human body part model may represent a non optimal match.
  • the colors may represent how tight or loose the PPE may be relative the human body part, such as for example red being shown on an area of the body part model where the PPE fits too tightly and green shown on an area of the body part model where the PPE fits too loosely.
  • the PPE selection system may output a comfort level based on a predetermined measurement scale, where the comfort level may reference a relative comfort of a PPE virtually worn by a user.
  • a comfort level may be determined by the amount of internal space measured between an inside part of a PPE and a corresponding body part.
  • a comfort level may be determined by a degree of permissible movement by a respective body part while a PPE is worn.
  • a comfort level may be determined for a respirator by determining whether the respirator maintains a seal with a facial area while the mouth of the user is being opened.
  • a user feeling may be determined by an objective comfort evaluation based on quantitative measurement.
  • a module may calculate a numeric pressure level upon the facial model as applied by the respirator model and compare the calculated pressure level with a set of
  • predetermined pressure ranges each associated with a specific comfort level are predetermined.
  • a PPE selection system may be used for predicting an optimal fitting PPE (e.g., respirator) for a specific user.
  • the system may includes an offline phase and a selection phase.
  • the offline phase may be performed during a fitting process of the PPE to the user in some exemplary embodiments. In other exemplary embodiments, the offline phase may be performed at some time prior to the fitting process of the PPE to the user.
  • one or more types of PPE may be selected to be associated with semantic information as in step.
  • several PPE are analyzed and processed to build a database of PPE having semantic properties.
  • the PPE types may include a variety of sizes and models of face respirators (e.g., masks).
  • a volunteer face model having semantic properties may then be correlated with the specific PPE such that the semantic properties of the face model are correlated with intersecting points or vertices of the specific PPE as depicted in step.
  • the volunteer face model may be chosen based on how well the specific PPE fits the volunteer face model. For example, if the specific PPE fits the volunteer face model well, then the respective volunteer face model may be used.
  • a specific PPE may be compared to a specific user model to determine a fit of the PPE with respect to a specific user model.
  • a region of interest to receive the PPE may be defined on the user.
  • the region of interest may be the face of the user.
  • the region of interest may be captured and modeled in a 3D format.
  • a scanning device or image capture device may scan the face of the user and form a 3D model of the user's face through one or a series of images.
  • a point cloud set may be defined on the 3D model of the user and semantic properties are applied to one or more of the defined points to generate a semantic face model as shown in step.
  • a database may be accessed to retrieve a specific PPE model to be compared to the specific user model, where the PPE model has semantic properties as shown by respirator model.
  • the database may be populated in the offline phase with a multitude of PPE models each representative of a specific model and size PPE and each having semantic properties.
  • the respirator model and the face model may be aligned to determine a fit.
  • the respirator model may be superimposed upon the face model.
  • the locations of the semantic properties of the respirator model and the face model may be then compared to determine a fit level of the respirator model to the face model as in step.
  • the distance between points on the respirator model and the user face model each having the same semantic properties may be determined for assessing a fit level of the respirator model to the user face model.
  • the respirator model having the best fit or highest fit level may be chosen for the specific user to wear as in step.
  • a final fit score determination process computes an overall fit score of PPE being virtually worn by a user, for example a 3D respirator model representative of a specific respirator being virtually worn by a 3D facial model representative of a specific user.
  • the determination process may include a graphical interface adapted for control by an administrator. For example, the administrator may individually control a variety of parameters to determine whether the 3D respirator model fits the 3D face model.
  • determination of whether the respirator model fits the 3D face model may be determined in response to a calculated estimated level of comfort parameter, an estimated level of face-seal parameter, and an amount of dead space parameter.
  • the estimated level of comfort parameter may be a degree of comfort felt by the user calculated by evaluating a contacting portion of the respirator model against the face model.
  • the estimated level of face-seal parameter may be a calculated gap between a seal of the respirator model and the face model.
  • the amount of dead-space parameter may be a calculated internal distance between the respirator model and the face model.
  • a final fit score may be calculated.
  • the final fit score may be representative of how well the respirator model fits the face model, or the perceived feeling or comfort of the user while wearing the respirator.
  • a final fit score of 100% may be representative of a perfect fit of the respirator model on the face model
  • a final fit score of 0% may be representative of a worst-case fit of the respirator model on the face model.
  • a final fit score of 75% may be representative of a very good fit of the respirator model on the face model and a final fit score of 25% may be representative of a below average fit of the respirator model on the face model.
  • each parameter may be weighted by a predetermined weighted function to arrive at a parameter result.
  • Each parameter result may be totaled to arrive at the final fit score.
  • the weighted function may be changed for a particular type of respirator or particular type of PPE.
  • the weighted function may be predetermined by the administrator or the user based on company or user preferences.
  • a 3D representation of the respirator model being worn by the face model may be illustrated in the graphical interface for display to the user.
  • a screenshot or printout of a PPE fit result may be provided to the user.
  • Some aspects of embodiments may be implemented as a computer system.
  • various implementations may include digital and/or analog circuitry, computer hardware, firmware, software, or combinations thereof.
  • Apparatus elements can be implemented in a computer program product tangibly embodied in an information carrier, e.g., in a machine-readable storage device, for execution by a programmable processor; and methods can be performed by a programmable processor executing a program of instructions to perform functions of various embodiments by operating on input data and generating an output.
  • Some embodiments can be implemented advantageously in one or more computer programs that are executable on a programmable system including at least one
  • a computer program is a set of instructions that can be used, directly or indirectly, in a computer to perform a certain activity or bring about a certain result.
  • a computer program can be written in any form of programming language, including compiled or interpreted languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, device driver, or other unit suitable for use in a computing environment.
  • Suitable processors for the execution of a program of instructions include, by way of example and not limitation, both general and special purpose microprocessors, which may include a single processor or one of multiple processors of any kind of computer. Generally, a processor will receive instructions and data from a read-only memory or a random access memory or both.
  • the essential elements of a computer are a processor for executing instructions and one or more memories for storing instructions and data.
  • Storage devices suitable for tangibly embodying computer program instructions and data include all forms of non-volatile memory, including, by way of example, semiconductor memory devices, such as EPROM, EEPROM, and flash memory devices; magnetic disks, such as internal hard disks and removable disks; magneto-optical disks; and, CD-ROM and DVD- ROM disks.
  • semiconductor memory devices such as EPROM, EEPROM, and flash memory devices
  • magnetic disks such as internal hard disks and removable disks
  • magneto-optical disks and, CD-ROM and DVD- ROM disks.
  • the processor and the memory can be supplemented by, or incorporated in, ASICs (application-specific integrated circuits).
  • the processor and the member can be supplemented by, or incorporated in hardware programmable devices, such as FPGAs and PLDs, for example.
  • each system may be programmed with the same or similar information and/or initialized with substantially identical information stored in volatile and/or non-volatile memory.
  • one data interface may be configured to perform auto configuration, auto download, and/or auto update functions when coupled to an appropriate host device, such as a desktop computer or a server.
  • one or more user- interface features may be custom configured to perform specific functions.
  • An exemplary embodiment may be implemented in a computer system that includes a graphical user interface and/or an Internet browser.
  • some implementations may be implemented on a computer having a display device, such as an LCD (liquid crystal display) monitor for displaying information to the user, a keyboard, and a pointing device, such as a mouse or a trackball by which the user can provide input to the computer.
  • a display device such as an LCD (liquid crystal display) monitor for displaying information to the user
  • a keyboard such as a keyboard
  • a pointing device such as a mouse or a trackball by which the user can provide input to the computer.
  • the system may communicate using suitable communication methods, equipment, and techniques.
  • the system may communicate with compatible devices (e.g., devices capable of transferring data to and/or from the system) using point-to-point communication in which a message is transported directly from the source to the receiver over a dedicated physical link (e.g., fiber optic link, point-to-point wiring, daisy-chain).
  • the components of the system may exchange information by any form or medium of analog or digital data communication, including packet-based messages on a communication network.
  • Examples of communication networks include, e.g., a LAN (local area network), a WAN (wide area network), MAN (metropolitan area network), wireless and/or optical networks, and the computers and networks forming the Internet.
  • implementations may transport messages by broadcasting to all or substantially all devices that are coupled together by a communication network, for example, by using omni-directional radio frequency (RF) signals.
  • Still other implementations may transport messages characterized by high directivity, such as RF signals transmitted using directional (i.e., narrow beam) antennas or infrared signals that may optionally be used with focusing optics.
  • RF radio frequency
  • USB 2.0 Firewire
  • ATA/IDE RS-232
  • RS-422 RS-485
  • 802.11 a/b/g/n Wi-Fi
  • Ethernet IrDA
  • FDDI fiber distributed data interface
  • token-ring networks or multiplexing techniques based on frequency, time, or code division.
  • Some implementations may optionally incorporate features such as error checking and correction (ECC) for data integrity, or security measures, such as encryption (e.g., WEP) and password protection.
  • ECC error checking and correction
  • WEP Secure Digital

Abstract

La présente invention concerne un appareil et des procédés associés relatifs à l'ajustement d'un masque virtuel sur un visage virtuel par un premier ajustement d'une région de menton du masque virtuel sur le visage virtuel, puis la détermination d'un angle de masque virtuel qui maintient la région de menton ajustée tout en ajustant simultanément une région de pont de nez du masque virtuel sur le visage virtuel, puis le calcul d'une métrique de qualité d'ajustement correspondant à la position ajustée. Dans un mode de réalisation illustratif, la région de menton ajustée peut comprendre la région de menton à forte courbure du menton. Dans certains exemples, un masque virtuel peut être appuyé virtuellement sur le visage virtuel à l'aide d'une force prédéfinie correspondant à une force d'un dispositif de fixation de masque d'un masque réel correspondant au masque virtuel. Dans un exemple de mode de réalisation, l'ajustement d'un masque virtuel sur un visage virtuel peut donner avantageusement une qualité d'ajustement de masque en peu de temps.
PCT/US2014/024098 2013-03-15 2014-03-12 Alignement de masque virtuel pour analyse d'ajustement WO2014150739A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US14/906,832 US10424113B2 (en) 2013-03-15 2014-03-12 Virtual mask alignment for fit analysis

Applications Claiming Priority (12)

Application Number Priority Date Filing Date Title
US13/839,186 2013-03-15
US13/839,056 2013-03-15
US13/839,056 US9361411B2 (en) 2013-03-15 2013-03-15 System and method for selecting a respirator
US13/839,186 US20140278320A1 (en) 2013-03-15 2013-03-15 System and method for selecting a respirator
US201361814905P 2013-04-23 2013-04-23
US201361814897P 2013-04-23 2013-04-23
US61/814,897 2013-04-23
US61/814,905 2013-04-23
US201361861294P 2013-08-01 2013-08-01
US61/861,294 2013-08-01
US201361917171P 2013-12-17 2013-12-17
US61/917,171 2013-12-17

Publications (1)

Publication Number Publication Date
WO2014150739A1 true WO2014150739A1 (fr) 2014-09-25

Family

ID=51580819

Family Applications (2)

Application Number Title Priority Date Filing Date
PCT/US2014/024200 WO2014150776A1 (fr) 2013-03-15 2014-03-12 Système d'ajustement de masque virtuel
PCT/US2014/024098 WO2014150739A1 (fr) 2013-03-15 2014-03-12 Alignement de masque virtuel pour analyse d'ajustement

Family Applications Before (1)

Application Number Title Priority Date Filing Date
PCT/US2014/024200 WO2014150776A1 (fr) 2013-03-15 2014-03-12 Système d'ajustement de masque virtuel

Country Status (1)

Country Link
WO (2) WO2014150776A1 (fr)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107256375A (zh) * 2017-01-11 2017-10-17 西南科技大学 一种电脑前人体坐姿监测方法
US11058838B2 (en) 2015-12-22 2021-07-13 Koninklijke Philips N.V. Customized mask with rigid support
WO2022069311A1 (fr) * 2020-09-30 2022-04-07 The University Of Birmingham Masque facial
EP3813960A4 (fr) * 2018-06-28 2022-04-27 The GMN Group LLC Dispositif et procédé d'ajustement de respirateur
WO2022106999A1 (fr) * 2020-11-19 2022-05-27 ResMed Pty Ltd Systèmes et procédés pour déterminer une rétroaction à un utilisateur en temps réel sur une vidéo en temps réel
US11417072B2 (en) 2018-05-21 2022-08-16 3M Innovative Properties Company Automated fitting of multiple articles of personal protection equipment concurrently worn by a user
US11474020B2 (en) 2017-09-01 2022-10-18 3M Innovative Properties Company Sensing element for respirator
US11534632B2 (en) 2017-09-01 2022-12-27 3M Innovative Properties Company Fit-test method for respirator with sensing system
US11748983B2 (en) 2018-05-21 2023-09-05 3M Innovative Properties Company Image-based personal protective equipment fit system using worker-specific fit test image data
US11793422B2 (en) 2017-09-01 2023-10-24 3M Innovative Properties Company Sensing system for respirator
CN117574466A (zh) * 2023-11-02 2024-02-20 齐鲁工业大学(山东省科学院) 一种基于面部三维数据聚类模型的呼吸面罩设计方法
US11918837B2 (en) 2017-09-01 2024-03-05 3M Innovative Properties Company Fit-test method for respirator with sensing system
CN111724439B (zh) * 2019-11-29 2024-05-17 中国科学院上海微系统与信息技术研究所 一种动态场景下的视觉定位方法及装置

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9361411B2 (en) 2013-03-15 2016-06-07 Honeywell International, Inc. System and method for selecting a respirator
WO2017112750A1 (fr) * 2015-12-21 2017-06-29 Nextteq Llc Joints de respirateur
US11113508B2 (en) 2018-06-28 2021-09-07 The Gmn Group Llc Personal protective equipment fitting device and method
GB2611553A (en) * 2021-10-07 2023-04-12 My Maks Fit Ltd Facemask

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020161416A1 (en) * 2001-04-26 2002-10-31 Justine Huang Nerves and muscle stimulator
US20060023228A1 (en) * 2004-06-10 2006-02-02 Geng Zheng J Custom fit facial, nasal, and nostril masks
US20060235877A1 (en) * 2004-06-04 2006-10-19 Ron Richard Mask fititng system and method
US20100283844A1 (en) * 2009-05-11 2010-11-11 Acep France Method and system for the on-line selection of a virtual eyeglass frame
US20120299945A1 (en) * 2006-05-05 2012-11-29 Parham Aarabi Method, system and computer program product for automatic and semi-automatic modificatoin of digital images of faces

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1961333A (zh) * 2004-02-12 2007-05-09 贝斯简·阿利万迪 从虚拟环境中制造商品的系统和方法
US8254637B2 (en) * 2006-07-27 2012-08-28 Resmed Limited Mask fitting system and method
US8194097B2 (en) * 2008-12-12 2012-06-05 Seiko Epson Corporation Virtual masking using rigid parametric modeling
US8327851B2 (en) * 2010-03-15 2012-12-11 Sleepnea Llc Respiratory mask with user interface

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020161416A1 (en) * 2001-04-26 2002-10-31 Justine Huang Nerves and muscle stimulator
US20060235877A1 (en) * 2004-06-04 2006-10-19 Ron Richard Mask fititng system and method
US20060023228A1 (en) * 2004-06-10 2006-02-02 Geng Zheng J Custom fit facial, nasal, and nostril masks
US20120299945A1 (en) * 2006-05-05 2012-11-29 Parham Aarabi Method, system and computer program product for automatic and semi-automatic modificatoin of digital images of faces
US20100283844A1 (en) * 2009-05-11 2010-11-11 Acep France Method and system for the on-line selection of a virtual eyeglass frame

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11058838B2 (en) 2015-12-22 2021-07-13 Koninklijke Philips N.V. Customized mask with rigid support
EP3393759B1 (fr) * 2015-12-22 2021-11-24 Koninklijke Philips N.V. Masque personnalisé à support rigide
CN107256375A (zh) * 2017-01-11 2017-10-17 西南科技大学 一种电脑前人体坐姿监测方法
US11474020B2 (en) 2017-09-01 2022-10-18 3M Innovative Properties Company Sensing element for respirator
US11918837B2 (en) 2017-09-01 2024-03-05 3M Innovative Properties Company Fit-test method for respirator with sensing system
US11793422B2 (en) 2017-09-01 2023-10-24 3M Innovative Properties Company Sensing system for respirator
US11534632B2 (en) 2017-09-01 2022-12-27 3M Innovative Properties Company Fit-test method for respirator with sensing system
US11748983B2 (en) 2018-05-21 2023-09-05 3M Innovative Properties Company Image-based personal protective equipment fit system using worker-specific fit test image data
US11417072B2 (en) 2018-05-21 2022-08-16 3M Innovative Properties Company Automated fitting of multiple articles of personal protection equipment concurrently worn by a user
EP3813960A4 (fr) * 2018-06-28 2022-04-27 The GMN Group LLC Dispositif et procédé d'ajustement de respirateur
CN111724439B (zh) * 2019-11-29 2024-05-17 中国科学院上海微系统与信息技术研究所 一种动态场景下的视觉定位方法及装置
WO2022069311A1 (fr) * 2020-09-30 2022-04-07 The University Of Birmingham Masque facial
WO2022106999A1 (fr) * 2020-11-19 2022-05-27 ResMed Pty Ltd Systèmes et procédés pour déterminer une rétroaction à un utilisateur en temps réel sur une vidéo en temps réel
CN117574466A (zh) * 2023-11-02 2024-02-20 齐鲁工业大学(山东省科学院) 一种基于面部三维数据聚类模型的呼吸面罩设计方法

Also Published As

Publication number Publication date
WO2014150776A1 (fr) 2014-09-25

Similar Documents

Publication Publication Date Title
US10424113B2 (en) Virtual mask alignment for fit analysis
WO2014150739A1 (fr) Alignement de masque virtuel pour analyse d'ajustement
US10061888B2 (en) System and method for selecting a respirator
Peruzzini et al. Exploring the potential of Operator 4.0 interface and monitoring
US20210104178A1 (en) System and method for three-dimensional augmented reality guidance for use of medical equipment
Plantard et al. Validation of an ergonomic assessment method using Kinect data in real workplace conditions
Endo et al. Dhaiba: development of virtual ergonomic assessment system with human models
Wu et al. Human-computer interaction based on machine vision of a smart assembly workbench
CN105190707B (zh) 基于三维建模的患者接口设备选择系统和方法
RU2677096C2 (ru) Способ и система выбора 3-d устройства интерфейса пациента
Wiedemann et al. Performance evaluation of joint angles obtained by the Kinect v2
CN113728394A (zh) 身体活动执行和训练的评分度量
CN105378802B (zh) 3d患者接口设备选择系统和方法
US20210322701A1 (en) Personal protective equipment fitting device and method
US20200050836A1 (en) Personal protective equipment fitting device and method
US10922899B2 (en) Method of interactive quantification of digitized 3D objects using an eye tracking camera
WO2022009041A1 (fr) Dispositif et procédé d'ajustement d'équipement de protection individuelle
Borduas et al. Reliability of mobile 3D scanning technologies for the customization of respiratory face masks
Kunz et al. Vision-based ergonomic and fatigue analyses for advanced manufacturing
Bonin et al. Digital Assessment of Anthropometric and Kinematic Parameters for the Individualization of Direct Human-Robot Collaborations
Wang Ergonomic-centric methods for workplace design in industrialized construction
Fang et al. An automatic method for computerized head and facial anthropometry
US11776212B1 (en) Anatomically conforming apparatuses, systems, and methods, and applications thereof
Schwarz-Muller Expedient methodology for capturing anthropometric data of encumbered personnel utilising 3D scanning technology
Krishnamurthy et al. Deriving statistical fit contours and shape of an aerosol mask from 3D head scans

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14769118

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 14906832

Country of ref document: US

122 Ep: pct application non-entry in european phase

Ref document number: 14769118

Country of ref document: EP

Kind code of ref document: A1