Patent Application
20190218847
The current disclosure concerns entrance management. More specifically, the current disclosure concerns machine vision systems and their application in passenger counting, automatic door activation and safety purposes.
Systems and methods are known in the art to detect the presence of a person or object at the entrance of an automatic door such as an elevator door or a pedestrian sliding door.
Example of such systems include infrared (IR) and/or microwave detector above the door which registers the increase of frequency of the emitted beam returning to the detector as a person moves towards the entrance, and more generally sense/calculate the motion coming from the objects to be detected.
Safety acoustic sensors are used for powered pedestrian doors, bus and train (transit) door safety, but lack a backup system and are prone to false alarm or lack of sensitivity due to unaccounted reflections and change in the speed of sound.
Furthermore, while an acoustic system can determine the distance to a target, it is ill suited to have sharp angle boundaries, lateral resolution and differentiate the nature of a detected target, whether it is a passenger or a suitcase or to determine the lateral position of the target. Also, due to pour spatial resolution, it does not differentiate well a target size. For example, a small object just under a sensor can give the same signal as a bigger object located at an angle, even though the distance is the same.
More recently, with the increasing power and speed of computer chips, 3D (three dimensional) cameras and other TOF (Time of Flight) based systems have begun to be used in detecting systems for automatic doors.
Examples of such systems are described in the following patent documents:
Drawbacks of current TOF based detecting systems include the large quantity of data to process, even in single function systems, resulting in either a relatively slow response of the system or the limited precision thereof. The use of a TOF based detection system for safety and/or at least two of safety, activation and/or people counting purposes increases the above-identified problem. Other known drawbacks of such an application include miscounting due to shade, other visual artifacts, and blindness.
The problem of the large quantity of data to consider in the use of a TOF camera detection system for an automatic door is solved by limiting the area of interest to zones which are parts of the field of view to a first zone for safety purpose and to a second zone for door activation purpose, and/or people counting, thus improving reliability and speed.
According to illustrative embodiments, there is provided a detecting system for an automatic door, the system comprising:
According to further illustrative embodiments, there is provided a detecting system for an automatic door, the system comprising a TOF camera having a field of view that is limited in a first zone for safety purpose and that is limited in a second zone for activation purpose.
The system includes one or a plurality of TOF cameras on a single side of the door or on both sides thereof. The use of a minimum of two cameras each on different sides of the door may require a reconstruction of the field of view (FOV) but results in i) a reduction of data processing due to the ability to ignore the door and ii) improved safety considering that the FOVs of both cameras overlap in the safety zone. It also allows to mount the camera above the door and therefore reduce the effect of the inherent dead zone for these systems.
The proposed systems can be used for example in powered pedestrian doors, bus, rail, etc.
According to some embodiments of the TOF based detection system, at least one ultrasound sensor system is further provided. The two technologies share hardware and information, resulting in improved reliability and safety and counting functions of the system, while reducing the amount of hardware used, space requirement and unit cost, and improving performance.
More specifically, in accordance to another illustrative embodiment, there is provided a detecting system for an automatic door, the system comprising:
Finally, in according to a further illustrative embodiment, there is provided a detecting system for an automatic door, the system comprising a TOF camera having a field of view that extends away from the TOF camera towards a floor under the automatic door within or near an entrance defined by the automatic door so as to define a light curtain; the light curtain having a configuration that is adaptable depending on predetermined criteria.
Other objects, advantages and features of illustrative embodiments of TOF based detecting system for an automatic door will become more apparent upon reading the following non-restrictive description of illustrated embodiments thereof, given by way of example only, with reference to the accompanying drawings.
In the appended drawings:
In the following description, similar features in the drawings have been given similar reference numerals, and in order not to weigh down the figures, some elements are not referred to in some figures if they were already identified in a precedent figure.
The use of the word “a” or “an” when used in conjunction with the term “comprising” in the claims and/or the specification may mean “one”, but it is also consistent with the meaning of “one or more”, “at least one”, and “one or more than one”. Similarly, the word “another” may mean at least a second or more.
As used in this specification and claim(s), the words “comprising” (and any form of comprising, such as “comprise” and “comprises”), “having” (and any form of having, such as “have” and “has”), “including” (and any form of including, such as “include” and “includes”) or “containing” (and any form of containing, such as “contain” and “contains”), are inclusive or open-ended and do not exclude additional, unrecited elements.
The present invention concerns a TOF based detecting system mounted to a door, to a door frame or adjacent thereto, to detect a body such as a person, a vehicle, parcels in or near the entrance or the door path defined by the door.
It is to be noted that the expression “body” is to be construed herein and in the appended claims as including a person, an animal, a limb and an object, such as, without limitation, a parcel and a bag or any part thereof.
The expression “door assembly” is to be construed herein and in the appended claims as including a door or a plurality of doors mounted to a door frame or more generally a door or a plurality of doors assembled together so as to selectively close an entrance.
The expression “door” is to be construed herein and in the appended claims as including any panel made from any material and having any shape and sizes which removably blocks an entrance.
The expression “automatic door” is to be construed herein and in the claim as a door that automatically opens and closes, for example through an electromechanical system when a person or object moves towards the entrance defined by the door.
The expression “door frame” is to be construed herein and in the appended claims as including any structural elements used to operatively mount the door, including a wall, a floor, a ceiling, a combination thereof and any assembly of beams, or mechanical parts.
The expression “target” is to be construed herein and in the appended claims as a body that enters a field of view and having characteristics that correspond to predetermined criteria for detection purposes.
A TOF (Time of Flight) based detecting system 10 for an automatic sliding door 12 according to a first illustrative embodiment will now be described with reference first to
According to the first illustrative embodiment, the TOF based detecting system 10 comprises two TOF cameras 14-14′, each one mounted on a door frame 16 above a door entrance 18 of the sliding door 12. The two 3D cameras 14-14′ have overlapping or physically separated field of views. An image processing algorithm is used to combine the images coming from both cameras 14-14′ into a single composite image. The resulting FOV 21 can be made as big as the sum of both cameras' field of views or smaller, wherein the safety zones provided by the two cameras FOV overlap and therefore provide to safe vital configuration.
Such a use of two TOF cameras allows reducing the door obstruction on the image while the door 12 closes and therefore reducing processing time and increasing reaction speed.
Since image processing algorithms are believed to be well-known in the art, they will not be described herein in more detail for concision purposes.
The TOF based detecting system 10 further comprises a controller 22 that is coupled to both TOF cameras 14-14′ so as to receive images therefrom in a continuous feed manner.
Each TOF camera 14 or 14′ includes several illuminating diodes and a 2D (two dimensional) receiver in the form of CCD (Charged Couple Device) cameras. According to another embodiment (not shown), the TOF cameras 14 are from another type than CCD cameras.
Since TOF cameras are believed to be well known in the art, they will not be described herein in more detail for concision purposes.
According to the illustrated embodiment, the TOF based detecting system
The safety zone 24, the two (2) door activation zones 26 and the counting zone 27 are defined by respective selected portions of the 3D images gathered by the two sensors 14-14′ and more specifically selected pixels therefrom. The whole of pixels in the 3D images from the two sensors 14-14′ is therefore not considered by the controller 22 in its analysis thereof.
More specifically, the controller 22 inspects first pixels of the 3D images that correspond to the safety zone 24 so as to determine information related to a safe use of the automatic door 12, and second pixels of the 3D images that correspond to the activation zones 26 so as to determine information related to the presence of person or target in view of activating the door 12. When a target is identified by one of the two activation zones 26, then the displacement of the target is followed in the counting zone to determine whether a number of persons crossing the door 12 should be incremented, decremented or stay the same.
The safety zone 24 is defined by a volume between the two sliding door panels 30 that extends within the full FOV 21 i) transversally a predetermined distance from the entrance 18, equidistant on both sides thereof, and ii) from a small distance ‘g’ from the floor 32, up to the upper limit of the full FOV 21.
The distancing of the activation zones 26 from the floor 32 allows ignoring the portions of the 3D images that correspond to the floor 32 so as to minimize the occurrence of false targets due to reflections from the floor 32. A distance ‘g’ as small as one pixel has been found to be sufficient to prevent reflections from the floor 32. The gap ‘g’ can also be of a different length than illustrated or omitted.
An image analysis is performed by the controller 22 to identify a target moving towards the defined safety zone 24 between the two panels 30. A conventional image data analysis software/algorithm is implemented on the controller 22 for such a task. Since such a data analysis is believed to be well known in the art, it will not be described herein in more detail for concision purposes. The safety analysis is not performed when the door 12 opens.
The controller 22 can be in the form of a solid-state device or a computer implementing the 3D data analysis functionalities required. The controller 22 is further configured for receiving and storing image data from the TOF cameras 14-14′.
As illustrated in
The position of the door panels 30 is determined either by a further image analysis in the full FOV 21 or by an encoder (not shown) provided on the door 12 and coupled to the controller 22.
The system controller 22 is further coupled to the door activation controller (not shown) or is further configured to act as a door activation controller to cause the door 12 to open when a target (not shown) enters the safety zone 26.
The same is true when a target enters either one of the two activation zones 26.
The activation zones 26 are positioned on each sides of the entrance 18, centered with the door entrance 18 and are each defined by a limited FOV of a respective camera 14 or 14′. Each activation zone 26 is generally in the form of a cone that extends within the overall FOV 21 from the cameras 14 or 14′ to the distance ‘g’ from the floor 32. The cones 26 are titled away from the door 12 so that the lowest section of each cone 26 is transversally distanced from the door 12.
The projection of the counting zone 27 on the lowest section of the full FOV 21 can be seen in
When a target (not shown) is detected by the controller 22 in the counting zone 27 on either side of the entrance 21, the displacement of the target is tracked to verify whether it crosses the entrance 21 and a number of persons/passengers crossing the entrance 21 is incremented or decremented, depending on the direction of displacement of the target and predetermined criteria. Such criteria include without limitations, the nature of the target as determined by the system 10.
The system 10 may be configured to trigger off the counting functionality and to ignore the counting zone 27, for example when the door 12 is closing, and then until the opening of the door 12 reaches a certain width.
Similar criteria can be used to trigger on or off the safety and activation zones 24 and 26 and functionalities.
Even though the system 10 according to the first illustrated embodiment includes two (2) activation zones, one on each sides of the entrance, a TOF based detecting system according to another embodiment may include a different number and different location of activation zones, including a single one on one side of the door 12.
Any one of the activation, counting and safety zones 26, 24 and 27 can be omitted and or modify. Also, other limited zones within the FOV of the cameras can be defined to detect targets corresponding to other predetermined criteria.
While the system 10 according to the first illustrative embodiment has been defined as being adaptive to the door panels position, such a feature can be omitted. Also, a TOF based detecting system according to another embodiment can be alternatively or additionally configured to be adaptive according to other criteria.
In some embodiments, the adaptive change in the configuration of the safety zone 24, activation zones 26 and/or counting zone 27 occurs on one side of the door 12 only or differently on both sides thereof.
For example, a TOF based detecting system for an automatic door according to another embodiment is configured to detect any target within either one of the detecting zones 24, 26 and 27 that appears to be fixed according to predetermined criteria. Such criteria can be based for example on the displacement (or lack thereof) of an identified target, its size, position within the zone, any combination thereof, etc. An example of such an identified target may correspond to a person (not shown) that stays nearly still near the door 12 for lack of movement opportunity, for example in a crowd near the entrance 18 defined by the door 12. Such an identified target becomes a fixed target and is ignored by the system 10 for safety, activation and/or counting purposes until it stops corresponding to the predetermined criteria.
As a further example, when a side of the entrance 18 is found by the system 10 to be crowded, the configuration and/or position of anyone of the safety 24, activation 26 and counting 27 zones can be changed to limit or prevent false detections.
Whether the adaptation concerns a fixed target, the position of the door panels or the like, it is achieved either by a change in the position or shape of the zone or the number of pixels constituting the latter, or any combination thereof.
A TOF based detecting system 40 according to a second illustrative embodiment is shown in
The system 40 comprises a single camera 14 that is mounted to the door frame 16 on one side of the door 12, resulting in a full FOV 41 in which the safety, activation and counting zones are selected.
According to this embodiment, some portions 42 in the FOV 41 on the side of the door 12 opposite the camera 12 becomes obscured by the door panels 30 as the door 12 closes. Such portions 42 are ignored by the system 40 in its image processing. Of course, the system 40 takes into account the door movements and change the active FOV as the door closes as described hereinabove.
As can be seen in
The creation of a crossing zone as illustrated in
It is to be noted that illustrative embodiments of a TOF based detecting system are not limited to any particular number of sensors 14-14′, number of zones and zone shape. According to some embodiments (not shown), a single sensor is provided for determining information related to each different aspect of a use of the automatic door, such as activation, safety, counting or else or a same sensor is used for all different aspects. Also, more sensors can be used to increase the resolution, coverage and detecting precision. Of course, the use of TOF cameras with a higher resolution will improve the detection precision, but may require more processing and cost.
As will become more apparent upon reading the following description with reference to further illustrative embodiments, a dual mode system can be provided, wherein one or more ultrasound sensors are included in addition to the TOF camera(s) 14-14′. Since such dual node system is similar to the TOF based system 10 described hereinabove, only the differences therebetween will be described herein for concision purposes.
A dual mode system 44 according to a fourth illustrative embodiment is shown in
While the dual mode system 44 is shown as having two TOF cameras 14 and 14′ such as described with reference to the first illustrative embodiment, it can be implemented with any number of TOF cameras 14. It is to be noted that the cameras can be either a TOF 3D system or a video camera.
According to still another embodiment, separate transducers are provided for the ultrasound emission and detection. Their numbers may also vary.
The system 44 acts on the principle that each of the TOF based and acoustic systems provides complementary information to the other. For example, according to some illustrated embodiments, the acoustic system precisely determines the distance based on reflections in a first spectral range and the TOF based system determines the type (human, non-human), and position of a target. Furthermore, the TOF based system may further create a light curtain configuration as described hereinabove.
For example, if the acoustic system informs the TOF based system of a target in the safety zone, the contrast/intensity is improved in the 3D images to better characterized such target. Also, if the acoustic system sees a false target but the TOF based system does not, then such target is determined by the controller to be static and part of the ‘missed learned’.
The acoustic system can be operated so that some of the detecting cones 50 are triggered on/off according to the position of the door panels 30.
The operation of such a dual mode system is illustrated in
It is to be noted that the expression ‘Machine Vision’ in
Similarly, the expression ‘acoustic system’ is to be construed in the description and in the claims as including at least one ultrasound transducer 48 and the controller 22, to which they are operatively coupled.
According to such an embodiment, the controller 22 is further used for driving the ultrasound transducers 48 and for receiving and analysing signals therefrom.
According to another embodiment, a further dedicated controller (not shown) coupled to the controller 22 is used to operate the ultrasound transducers 48 and/or for analysing signals therefrom.
Since the operation of an ultrasound transducer is believed to be well known in the art, it will not be described herein in more detail for concision purposes.
During the counting process, the machine vision system uses information gathered by the acoustic system to confirm or determine that there are people going in the counting zone 27. This allows eliminating under counting due to blindness, over counting due to shades or other vision artefacts, and absorption of the signal within the machine vision operational spectrum.
The system may be further configured so that an identification or confirmation by the acoustic system that there are people in the counting zone 27 triggers a change in contrast and or sensitivity of the vision system camera to allow it to detect these persons. If the vision system is still unable to detect a target, then the acoustic detection is adapted and used to verify the legitimacy of the initially detected target.
According to a more specific embodiment, the analysis is such as to allow demographically discriminating detected humans or animals.
According to another embodiment, the vision system additionally or alternatively characterizes the movement of the target and whether it is going towards the safety area. The acoustic system then determines if the object has effectively passed through the counting zone 27, using criteria such as distance and angle. The combined information increases the people counting accuracy.
According to the fourth illustrative embodiment, both the acoustic and machine vision systems can confirm to the other that there is a passenger in a safety zone (see
Detection of a new target by the acoustic system is based on distance, signal strength and size, or the absence of signal. The machine vision system determines whether it can detect an object (which may be a small object) and, if so, the type of object. Using this information, control of the door 12 (opening/closing/no change) can be made more reliably.
According to another embodiment, each of the TOF based and acoustic system obtains its results independently and either one or a third unit is used to handle the analysis of both system results in order to trigger an action accordingly.
The number and positions of the TOF cameras 14-14′ and ultrasound transducers 38 may vary. They can be mounted to the door frame 16, to the door panels 30 and to any other structure nearby having a line of sight to the door 12 or to a region adjacent thereto.
As mentioned hereinabove, the TOF cameras 14-14′ and ultrasound transducers 38 are coupled to one or more controllers to provide the above-described functionalities. Such controller is coupled directly or indirectly to the automatic door controller or includes the automatic door control functionalities.
According to a further illustrative embodiment, the acoustic system is configured with adaptive quality i) by triggering on and off selected ultrasound sensors from a sensing assembly depending on their relative position to the door or the relative position of their beam; and/or ii) by modifying the shape and/or size of selected beams.
The acoustic system may further use an adaptive threshold technology, such as the one described in U.S. Pat. No. 7,130,244 B2 issued on Oct. 31, 2006 to Gal et al., that allows adapting the detection to changes in the detecting system environment or an adaptive dead zone technology that modifies the detection zones based on the door position, such as the one described in the U.S. Pat. No. 6,344,642 issued on Feb. 5, 2002 to Agam et al. Furthermore, the system can learn and recognize the door by its acoustic reflection pattern and update the detecting zone accordingly.
Any other type of sensor that is configured to operate outside of the visual spectrum can also be used in a dual mode system in addition to a TOF based detecting sensor.
It is to be noted that many other modifications could be made to the embodiments of an TOF based detecting system for an automatic door as described hereinabove. For example:
Even though a TOF based detecting system for an automatic door has been described hereinabove by way of illustrated embodiments thereof, it can be modified. It is therefore to be understood that numerous modifications may be made to the illustrative embodiments and that the scope of the claims should not be limited by the preferred embodiment, but should be given the broadest interpretation consistent with the description as a whole.
This application is a continuation of International Patent Application No. PCT/CA2017/000213, filed on Sep. 29, 2017, which takes priority from U.S. Provisional Patent Application No. 62/403,631, filed Oct. 3, 2016, the contents of each of which are incorporated by reference herein.
| Number | Date | Country | |
|---|---|---|---|
| 62403631 | Oct 2016 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | PCT/CA2017/000213 | Sep 2017 | US |
| Child | 16367507 | US |
