Object direction detection system

Abstract

To provide a system for tracking a face direction and a future face rotation of a guest who rides on a carnival ride in a pleasure ground.SOLUTION: A face direction system of a carnival ride comprises one or more sensors and a controller. The controller receives carnival ride cart data which indicates presence and a position of a carnival ride cart on a carnival ride, from the one or more sensors, determines a direction of a carnival ride cart direction on the basis of the carnival ride cart data, then receives from the one or more sensors, position data of a guest which indicates a position of the body of a guest, a position of the face of the guest or a position of a combination of the body and face of the guest, then determines the face direction, face rotation of the guest or combination thereof on the basis of the carnival ride cart direction and position data. Then, on the basis of the determined face direction, the face rotation or the combination of them, the controller sends the data indicating the determined face direction, the rotation of face or the combination of them, to a downstream controller for subsequent control.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] FIELD OF THE DISCLOSURE The present disclosure relates generally to the field of camera-based object detection. In particular, embodiments of the present disclosure relate to a system and method for identifying object orientation using a camera. [Background technology]

[0002] In many cases, it may be useful to understand the orientation of an object. Traditionally, wearable devices or other known object features have been used to help facilitate object orientation detection. However, providing and using wearable devices can be costly and cumbersome. Furthermore, the known features may vary from object to object and therefore may be difficult to detect and use as a reference.

[0003] For example, many amusement park type rides include ride vehicles that carry passengers along a ride path defined, for example, by a track. The ride path may include multiple features, including tunnels, turns, uphill, downhill, and loops, throughout the course of the ride. The ride path may also include show scene features, such as lighting effects, animated or robotic figures, that may provide supplemental content and / or enhance the immersive experience of the guest. However, due to the complexity of some amusement rides, determining the body and / or facial orientation of a guest within the ride, such as to activate or deactivate features when the guest is viewing a show scene feature along the ride path, may be difficult without the use of guest wearable ride gear. However, providing wearable ride gear may be costly and require additional time to implement as opposed to an amusement park type ride that does not use such gear. Summary of the Invention

[0004] The following summarizes some embodiments common in scope to the subject matter of the original claims. These embodiments are not intended to limit the scope of the disclosure, but rather to provide merely a brief summary of some disclosed embodiments. Indeed, the disclosure may include a variety of forms that may be similar to or different from the embodiments set forth below.

[0005] The embodiments described herein relate to a system that can track the face orientation and future face rotation of a guest on an amusement ride. As used herein, the term "face orientation" refers to the guest's head position with respect to a particular rotation angle. The term "face rotation" refers to the actual movement or range of movement of the head to reach a particular rotation angle. In some embodiments, the system can determine the face orientation and face rotation using a combination of ride cart orientation on the ride and skeletal tracking. In other embodiments, the system can determine facial features along with the face orientation and face rotation using face and / or body image detection. Additionally, machine learning can be used in determining the face orientation based on repeated determinations.

[0006] Although the following description focuses primarily on amusement park type rides, this is not intended to limit the scope of the invention, and in fact the object orientation detection system can be used in a number of different applications that are completely separate from amusement park type rides.

[0007] In one embodiment, the vehicle face direction system includes one or more sensors and a controller, the controller receives ride cart data from the one or more sensors indicative of a presence and position of a ride cart on the vehicle, determines a ride cart orientation based on the ride cart data, receives guest position data from the one or more sensors indicative of a position of the guest's body, face or combination thereof, determines a guest's face direction, face rotation or combination thereof based on the ride cart orientation and position data, and transmits data indicative of the determined face direction, face rotation or combination thereof to a downstream controller for subsequent control.

[0008] In one embodiment, the method includes receiving ride cart data from one or more sensors indicating a presence and position of a ride cart on a ride, determining a ride cart orientation based on the ride cart data, receiving guest position data from the one or more sensors indicating a position of the guest's body, face or a combination thereof, determining a guest's face orientation, face rotation or a combination thereof based on the ride cart orientation and position data, and transmitting data indicative of the determined face orientation, face rotation or combination thereof to a downstream controller for subsequent control.

[0009] In one embodiment, the tangible, non-transitory machine-readable medium includes machine-readable instructions that, when executed by one or more processors of the machine, cause the machine to: receive ride cart data from one or more sensors indicating a presence and position of a ride cart on a ride; determine a ride cart orientation based on the ride cart data; receive guest position data from the one or more sensors indicating a position of the guest's body, face or a combination thereof; determine a guest's facial orientation, facial rotation or a combination thereof based on the ride cart orientation and position data; and transmit data indicative of the determined facial orientation, facial rotation or combination thereof to a downstream controller for subsequent control. [Brief description of the drawings]

[0010] These and other features, aspects, and advantages of the present disclosure will be better understood from the following detailed description when read in conjunction with the accompanying drawings, in which like elements are designated with like reference characters throughout.

[0011] [Figure 1] FIG. 1 illustrates a block diagram of an amusement park ride that includes an attraction having show scenes along the ride path according to an embodiment of the present disclosure.

[0012] [Diagram 2] FIG. 1 is a schematic diagram of a face direction system including a show scene controller used to control show scene features, according to an embodiment.

[0013] [Diagram 3] 3 is a block diagram of the face direction system of FIG. 2 that uses skeleton tracking to determine face direction and then uses this to control an animated figure, according to one embodiment.

[0014] [Figure 4] FIG. 13 is a process flow diagram for determining a guest's face direction according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] One or more specific embodiments of the present disclosure will be described below. In order to concisely describe these embodiments, not all of the features of the implementations may be described herein. It should be understood that in the development of any such implementation, as in any engineering or design project, numerous implementation-specific decisions must be made to achieve the developer's particular objectives, such as compliance with system-related and business-related constraints that may vary from implementation to implementation. Moreover, it should be understood that such development efforts may be complex and time-consuming, but are routine undertakings of design, fabrication and manufacture for those skilled in the art having the benefit of this disclosure.

[0016] When introducing elements of various embodiments of the disclosure, the articles "a," "an," and "the" are intended to mean the presence of one or more of the element. The terms "comprising," "including," and "having" are intended to be inclusive and mean that there may be additional elements other than the listed elements. It should also be understood that references to "an embodiment" or "an embodiment" of the disclosure are not intended to be interpreted as excluding the existence of additional embodiments that also contain the recited features.

[0017] Amusement parks feature a wide range of entertainment, such as amusement rides, live shows, and games. Different types of entertainment can include show scene features, such as themed scenes along the ride path, that enhance a guest's experience at the amusement park. Show scene features can include effects, objects, or equipment along the ride path that are activated while a guest is viewing a particular show scene. For example, a robotic figure along the ride path can include elements or objects that change or activate based on the detection of the presence of the cart and / or guest along the ride path. Show scene changes can include animated equipment fluctuations, and lighting, changes in the position of the ride cart, etc.

[0018] Activation of typical show scene features (e.g., lighting, animation) may be triggered by limited information about the attraction and / or about the guest viewing the show scene objects along the ride path. For example, the trigger may be based on the location of the ride cart and / or the guest. Thus, show scene features provided on the ride path may be turned on regardless of the guest's location within the ride cart. This may result in a less personalized experience typically achieved for each iteration of the attraction.

[0019] However, it may be desirable to provide more granular triggers that result in more customized attraction changes based on specific characteristics of the attraction guest. Provided herein is a technique that facilitates the determination of a guest's face direction and face turning while riding the ride without the use of face or head gear for the guest while riding the ride. This determination of the guest's face direction and face turning can be used to trigger or initialize show scene features. For example, a show scene feature, such as an animation or special effect, can be provided in response to the guest's face being oriented to be within the viewing range of a particular object or facility on the ride path that provides the show scene feature. Generally, the ride cart travels along the ride path with the guest's body or head moving and turning to view features along the ride path. Effects and animations can be triggered to initialize in response to detection of an active ride cart on the ride. However, providing accurate guest face direction and face turning at a particular location on the ride path, such as within the viewing range of an object or facility that provides a special effect or animation on the ride, can be complicated.

[0020] It should be noted that although the examples provided herein may be generally illustrated in the context of amusement parks and ride attractions, such as using the present guest face direction determination technology to facilitate activation of features in show scenes on a ride route, the technology in the present disclosure may also be applied to other non-amusement park related conditions and / or contexts. Thus, it should be understood that the present examples merely reflect an example of a face direction system on a ride to provide an illustrative context, and should not be considered as limiting the applicability of the present approach. Rather, it should be understood that the present approach is applicable to any situation in which the precise direction of people or objects moving (e.g., turning their faces) within an environment can be determined for business, entertainment, or other purposes.

[0021] With this in mind, Figure 1 is a block diagram of an amusement park ride 100 that can operate with the disclosed facial direction system, according to an embodiment. Specifically, the amusement park ride 100 can include one or more ride carts 112, with or without guests, and show scene features along a ride path 108, including animated figures 102, lighting effects 104, and displays 106. The show scene features can be operated to change based on the orientation of the ride cart 112 and the detected facial direction and / or head turning of the riding guests.

[0022] As shown, show scene features such as the animated figures 102, lighting effects 104 and displays 106 can be controlled to react via a show scene controller 110. In an embodiment, the show scene controller 110 of the face direction system can be centralized to control one or more show scene features and synchronized with at least one or more show scene features located throughout the ride path 108 of the amusement ride 100. The show scene controller 110 receives contextual position information (e.g., ride cart 112 position data, as well as guest body and / or head position data) via one or more sensors and / or processors located on or around the show scene features, the ride cart 112 and / or the ride path 108. The face direction system can use this information to determine a face direction as described in detail in FIG. 2.

[0023] Once the face direction is determined, a control signal can be sent to one or more systems controlling the show scene features to trigger one or more of the show scene features based on the face direction. Contextual position information (e.g., sensory data indicative of the position of the ride cart 112 on the ride track 116 as well as guest body and / or head position) can also be sent to the monitoring system 111. The monitoring system 111 can be a management system that not only monitors show scene changes such as animation or lighting control signals sent from the show scene controller 110 in response to the direction and face direction of the ride cart 112, but can also control or reconfigure the show scene controller 110. In this manner, the monitoring system 111 can reset show scene feature changes or disable animations and effects determined by the show scene controller 110. For example, as described in detail in FIG. 2, the show scene controller 110 can automatically send animations to the animated figure 102 in response to a guest looking towards the animated figure 102. Similarly, if the ride operator 122 notices any issues with the animated figure 102, the ride operator 122 can manually stop the animation via a stop mechanism (e.g., a button) that disables the animation beyond a decision by the show scene controller 110. The monitoring system 111 can also be used to recalibrate sensors on or around the show scene. Additionally, the monitoring system 111 can also be used to reset the face direction determination algorithm of the show scene controller 110. In some embodiments, the monitoring system 111 and the show scene controller 110 can be implemented as a single controller.

[0024] As described above, the show scene controller 110 of the face direction system can utilize a combination of sensor data received by sensors on or near the show scene features, data from the ride cart 112 and / or the ride path 108 to detect the position of the ride cart 112 and the guest's body / head, and further determine the face direction, alignment, and / or estimate the face rotation. This determination can be used to control the animations and effects generated throughout the show scene accordingly. Based on the orientation of the ride cart 112, a 3D map of the guest on the ride can be generated. The 3D map generation can be based on the known position and orientation of the ride trajectory 116, as well as known features within the ride 100 (e.g., the ride trajectory 116 turns toward the show scene feature at a midpoint on the ride 100). For example, the 3D map can generate the orientation of the ride cart 112 based on the position of the ride trajectory 116. Additionally, the guest position data in the 3D map can be based on the orientation of the ride cart 112 and the known ride cart 112 configuration (e.g., the seats face forward as the ride cart 112 moves along the ride path 108). This 3D map can be used to recreate the 3D position of the guest, and then the face orientation can be accurately determined, particularly with respect to known show scene features in the ride path 108. The 3D map can also determine where the head may turn, such as to face toward a show scene feature that the guest is expected to view based on the determined face orientation. Additionally, machine learning can be used to accurately predict face positions based on repeated determinations made using the sensor data. Machine learning can also be further used to predict face orientation based on learned guest face orientations during specific points on the ride path 108 (e.g., that faces are often oriented toward a specific show scene feature).

[0025] As shown, the ride cart 112 can be detected using radio frequency identification (RFID) tags 118 integrated into the ride cart 112 or guest wearable devices, weight sensors 120 along the ride track 116, and / or infrared (IR) markers 119 located on the ride cart 112. The RFID tags 118 can communicate with electronic readers 123 integrated into different locations on the ride 100, such as on the ride track 116, to indicate the presence of the tag. Thus, the electronic readers 123 located on the ride path 108 can scan the RFID tags 118 on the ride cart 112 as the ride cart 112 passes the electronic readers 123. Additionally, the RFID tags 118 can be located in specific areas of the ride cart 112 and / or can be different types of RFID tags 118. In this manner, the location and / or type of RFID tags 118 can be used to determine the front and rear of the ride cart 112. This information can be used to more accurately predict the guest's location within the ride cart 112, such that the guest is predicted to be facing forward of the ride cart 112. Similarly, IR markers 119 can also be placed on the ride cart 112. The IR markers 119 can include IR reflective surfaces that are not optically visible but can be detected by one or more cameras 114. The cameras 114 can be equipped to detect IR radiation reflected from the markers. Additionally, RFID tags or IR markers can also be placed periodically on the ride track 116 or ride path 108 at specific intervals or at specific points of interest on the track. The cameras 114 that can be used to detect the body and / or head position of the cart 112 and / or the guest can be configured to conserve power by activating after the ride cart 112 approaches the viewing range of a particular camera. Similarly, tracking of the ride cart 112 at specific points on the ride track 116 can be used to activate or deactivate show scene features along the ride path 108. As an example, the RFID tag 118 can indicate that the ride cart 112 is at the beginning of its journey and therefore turn off show scene features located towards the end of the journey.When the ride cart 112 approaches a point of interest on the ride 100, such as a predetermined distance (e.g., 3 meters) to a particular show scene feature, which can be determined using a proximity sensor 121 or based on a 3D map, the show scene controller 110 can activate such feature accordingly. Additionally or alternatively, an RFID tag 118 or IR marker 119 can be positioned to identify the location of the ride cart 112 on the track (e.g., a straight track or a turning section).

[0026] Additionally, the guest wearable device may also include an RFID tag 118 or an IR marker 119 that can be used to indicate the presence of the guest in the ride cart 112 and further trigger a face direction determination. As an example, an electronic reader 123 for the guest wearable device may be placed on the seat of the ride cart 112 to indicate the presence of the guest when scanned. Additionally or alternatively, the IR marker 119 on the guest wearable device may also be detected by a camera using the techniques described above.

[0027] Weight sensors 120 may also be mounted on the ride track 116 to indicate the presence of the ride cart 112 on the ride track 116 based on exceeding a predetermined weight threshold. After locating the ride cart 112 using the techniques described above (e.g., RFID tag 118, IR marker 119) to generate a 3D map, the camera 114 may observe an area of ​​interest, such as the position and / or location of the ride cart 112 within the ride path 108. Thus, the capture of guest body and / or head position data may be limited to the guest on the ride cart 112. Furthermore, since the ride operator 122 knows the seat location of the ride cart 112, the camera 114 may be configured to narrow the observation range to focus on a specific area around the seat. Thus, the location of the ride cart 112 and other sensor mechanisms described herein may be used to locate the guest and accurately determine the face direction. Additionally, the face direction may be used to determine the observation direction within the ride path 108.

[0028] As shown, one or a series of cameras 114 may be positioned throughout the amusement ride 100 along the ride path 108 to determine the body and / or head orientation of one or more guests. Although a series of cameras 114 are shown and described herein to detect guest orientation, a single camera 114 may also be used. Furthermore, in some embodiments, a single camera 114 may be used in addition to or instead of the techniques described above to identify the location and orientation of the ride cart 112, which is used to generate the 3D map.

[0029] The camera 114 can be set at a predetermined distance from the show scene features, such as the animated figure 102, the lighting effects 104, and the display 106. The predetermined distance can be a distance that can allow for the capture of the head and at least the upper body of the guest to allow accurate determination of head direction (e.g., front head, back head facing the show scene feature), etc., depending on the show scene feature. For example, the camera 114 can observe a 3 meter circumference of the animated figure 102. In this way, any guest on the ride cart 112 close to the animated figure 102 can be easily detected. Additionally or alternatively, the camera 114 can be incorporated into a show scene feature, such as the body of the animated figure 102. The incorporation can be positioned to sense a clear view of one or more guests, regardless of whether the guest is located in front of the animated figure 102 in the ride cart 112 or has passed the animated figure 102. For example, the camera 114 can be incorporated into the eyes of the animated figure 102 to detect a guest in front of the animated figure 102. Additionally or alternatively, the predetermined distance may not be uniform for all show scene features depending on the desired effect on the guest. As an example, the lighting effects 104 along the ride path 108 may be switched on before the ride cart 112 and / or the guest are in front of it. Thus, one or more cameras 114 may determine guest presence at a distance greater than the distance of the animated figure 102. Additionally, the cameras 114 may be positioned to exclude a field of view that may not include guests of interest. For example, as shown, the ride operator 122 may operate the ride such that guests waiting in the queue 107 of the ride 100 are signaled to board the ride cart 112. Because the ride operator 122 and guests in the queue 107 are not on the ride 100 or in the ride cart 112, their face directions relative to the show scene are not of interest and therefore may be excluded from the observation range of the camera 114.

[0030] The camera 114 may also be used to capture data used as input to one or more processes of the face direction system described herein. In one example, the show scene controller 110 may use the data from the camera 114 to detect a human body. That is, in addition to or instead of generating a 3D map of the ride cart 112, the presence of the captured image may be used to determine the presence and location of any guest on the ride cart 112 and further determine the face direction. Skeleton tracking may be used at least in part to further determine the body and / or face location of the guest on the ride cart 112. Skeleton tracking may include using real-time or captured images of the guest on the ride cart 112 by the camera 114 and comparing these images with a recognized skeletal model indicative of human body skeleton detection. For example, the skeletal model may include the relative positions of important joints, bones and limbs of the human body skeleton. Joints located close to each other on the skeletal model may indicate that the body and / or face is oriented in a particular direction. Additionally, the skeletal model may also be used to predict the future face direction of the guest based on the current face and body position. For example, if the skeletal model indicates that the face orientation is forward but the body is slightly turned to the right (e.g., shoulder joints are closer to each other toward the right end of the skeletal body) based on the body and face joint positions, then the face orientation can be predicted to be right as well. Additionally, ride cart 112 data indicating specific show scene features at multiple points along the ride path 108 can also be used in conjunction with skeletal tracking to make future face rotation predictions (e.g., that the face is likely to turn toward the known show scene feature). Thus, comparison of the captured images of the guest to the skeletal model can be used for guest detection and body / face orientation. In some embodiments, the range of face rotation can be determined based on the ride cart 112 orientation and / or body orientation. For example, a typical guest can rotate his / her head by a maximum head rotation (e.g., 90 degrees) in either direction relative to a known cardinal direction.Assuming the user is facing forward in the ride cart 112, the orientation of the ride cart 112 can be used as the known base orientation to determine the range of facial rotation. Additionally or alternatively, if the guest's body orientation is known, it can be used as the known base orientation to determine the range of facial orientation. As an example, if the orientation of the ride cart 112 is used as the known base orientation and the orientation of the ride cart 112 is 45 degrees, the range of the guest's facial rotation can be determined as the known base orientation plus or minus a maximum head rotation. Here, the range of the guest's facial rotation is 45 degrees plus or minus 90 degrees, thus ranging from -45 to 135 degrees. Similar calculations can also be performed using the guest's body orientation as the known base orientation.

[0031] In some embodiments, the fast paced ride 100 may use less granular face orientation analysis and may use skeletal tracking and ride cart 112 data to determine face orientation. As an example, ride cart 112 orientation determination using the techniques described above (e.g., known front and rear of the ride cart 112, ride path 108, ride trajectory 116, show scene features 110 on the ride path 108) may indicate an estimated body position of the guest based on the orientation of the ride cart 112 and any show scene features along the ride path 108 that the guest can see. Thus, the ride cart 112 data may indicate the general location of the guest within the ride cart 112 and its relative position to show scene features that may be located along the ride path 108. Additionally, the skeletal model from skeletal tracking may further provide accurate body and face orientation of the guest. Positions of joints closer to one end of the skeletal model from each other may indicate that the guest is facing the same direction as the show scene features along the ride path 108. Therefore, face direction determination can be performed using a combination of data from the vehicle cart 112 and skeleton tracking.

[0032] On the other hand, a slower paced vehicle 100 may use more granular face orientation analysis, including facial feature analysis, etc. However, the skeletal image provided by skeletal tracking may not show exact facial features. Thus, the skeletal image data may be further processed to identify faces using facial recognition and / or body shapes using body recognition.

[0033] In some embodiments, such as the slow-paced ride 100, facial image detection may consider multiple facial features together to determine that a face has been accurately detected. The features considered may be facial points of interest and their intersections (e.g., two eyes aligned along a horizontal line very close to a nose along a vertical line detected at or near the point between the eyes). Furthermore, the camera 114 may be used for full body detection or upper body detection depending on the observation range of the camera 114, such that the guest direction (e.g., the guest is facing forward) may be accurately identified using the body features. For example, the body detection features may consider the body features such as the face, arms, and legs and their intersections. The body detection may further indicate the guest's body orientation to predict the guest's facial direction and / or future facial direction. As an example, if the guest's face is detected as facing forward, but the body is detected as slightly turned to the right, the facial direction may also be predicted to be facing right. Additionally or alternatively, ride cart 112 data indicating the position of the ride cart 112 on the ride path 108 and relative position to show scene features may be used in conjunction with body and / or face image detection to predict future face rotation directions. Additionally, the camera 114 may be used for thermal recognition to detect body heat and may be used to more accurately predict guest presence and orientation. Additionally, the camera 114 may include infrared (IR) capabilities to provide night vision capabilities to enable day and night guest detection. As discussed above, IR capabilities may also be used for IR markers 119 to detect the ride cart 112.

[0034] Some of the different types of sensors, such as RFID tags 118, IR markers 119 and / or camera 114 inputs used to detect the location of the ride cart 112 in the ride 100 to generate the 3D map, may be used in conjunction with further camera 114 inputs used to determine the guest's body and / or face, such as by skeletal tracking. Thus, based on the detection of the ride cart 112 being present at a particular location in the ride 100, the camera 114 may be initialized for skeletal tracking, body detection and / or face detection to determine the face direction. In some embodiments, after determining the guest's face direction, a show scene feature may be initialized to provide entertainment to guests facing a feature within a predetermined range. That is, an animation or effect may be triggered if the sensed distance is below a threshold and / or if the guest is determined to be within the virtual field of view of the show scene feature and facing the feature.

[0035] As described above, after determining the location of the ride cart 112 along with the guest's body and / or head position used to determine the guest's facial orientation (e.g., that the face has rotated right relative to the body), the algorithm of the face direction system can be modified based on machine learning. Specifically, the face direction determination algorithm can be changed in response to ongoing or periodic determinations. As an example, if the guest's facial orientation is determined to be positioned toward a show scene feature at a particular point on the ride path 108, the show scene controller 110 can learn to predict the facial orientation at the same location via machine learning. Similarly, if the facial orientation is determined to be forward facing at a particular point on the ride path 108, the show scene controller 110 can make a similar prediction using machine learning. Additionally, when the face is determined to have rotated at a particular point on the ride path 108, machine learning can also be used to predict the facial rotation. Thus, machine learning can be used to identify patterns of facial orientation and future rotations such that the show scene controller 110 makes accurate facial orientation determinations and predictions. According to some aspects described herein, machine learning methods that can be used to make predictions regarding face detection and orientation, such as expected face turning, can further be used to accommodate show scene animations or effects being performed. The machine learning can be based on face orientation determination using detection mechanisms (e.g., IR markers 119 to detect the ride cart 112, skeletal tracking, and / or facial recognition to detect faces), update or modify face orientation algorithms based on the latest data through detection and reconfiguration execution or other suitable techniques.

[0036] As mentioned above, the face direction system's show scene controller 110 can be sent a combination of detected sensor data indicating the direction of the ride cart 112 and / or the guest's location within the 3D map of the ride 100. As a result, the face direction system can determine the change in any of the show scene features after using these sensor data to determine the face direction. To illustrate, FIG. 2 is a block diagram of a face direction system 150 having a show scene controller 110 used to control a show scene feature 160, which can include a display 106, lighting effects 104, and an animated figure 102. The show scene controller 110 (e.g., an electronic controller) can include a memory 156, a processor 154, a communication circuit 158, and an input device 152. In some embodiments, the memory 156 is configured to store instructions, data, and / or information, such as a database of images, animations, or effects for the show scene features. Additionally or alternatively, the memory 156 can also include a user-defined algorithm for generating a 3D map of the ride cart within the ride and an algorithm for determining the face direction. It should be noted that while the examples provided herein may explicitly depict and describe show scene features as being limited to displays 106, lighting effects 104, and / or animated figures 102, the disclosed techniques may also include using face direction to control other show scene features or effects on the vehicle path 108, such as changing the path of the vehicle trajectory 116, figures spraying water, and the like.

[0037] As described in further detail below, in some embodiments, the processor 154 is configured to receive input signals from the input device 152 related to detection of the position of the ride cart 112 and / or the guest's orientation from one or more sensors as described above, and use these to determine a face direction using the techniques described herein. Images, effects, or animations from a database can then be provided based on the face direction and the guest position within the ride determined according to the position of the ride cart 112. In this manner, the show scene controller 110 can determine the face direction and further control the show scene feature accordingly when the guest is within a predetermined range of a particular show scene feature and / or when the guest's face is determined to be oriented toward a particular show scene feature. As described above, sensor data can be provided to the input device 152 by the camera(s) 114, weight sensor(s) 120, IR marker 119, RFID tag 118, and / or proximity sensor(s) 121. It should be understood that the illustrated system is intended to be illustrative only, and that certain features and components may be excluded and various other features and components may be added for ease of implementation in accordance with the disclosed embodiments.

[0038] The show scene controller 110 can facilitate control of the show scene features 160, which uses the processor 154 to select images to display, animations to perform, and special effects that can be stored in a database in the memory 156 based on the sensor data received on the input device 152. The database can include a list of available images, such as images related to the theme of the ride. As an example, the ride 100 of FIG. 1 depicts a circus theme, and thus a circus image can be displayed on the display 106 when the guest's face is facing the display 106. The database can also include animations for the animated figures 102 (e.g., clowns) to perform. Continuing with the circus theme example, the animations can include, but are not limited to, juggling, smiling and waving, saying hello, and the like. Additionally, the lighting effects 104 options can include colored lighting, strobe lighting, shadow lighting, and / or lighting that creates circus-related images that are displayed throughout the ride path 108 (e.g., tunnel walls on the ride path 108).

[0039] In the illustrated embodiment, the processor 154 of the show scene controller 110 can execute instructions stored in the memory 156 to perform operations such as generating a 3D map, determining a face orientation based on sensor data received at the input device 152, and selecting or modifying a show scene feature 160 (e.g., causing the animated figure 102 to juggle when a detected face is directed toward the animated figure 102). Thus, in some embodiments, the processor 154 can be one or more general-purpose microprocessors, one or more application specific processors (ASICs), one or more field programmable gate arrays (FPGAs), or any combination thereof. Also, the memory 156 can be a tangible, non-transitory computer-readable medium that stores instructions executable by the processor 154 and sensor data to be processed by the processor 154. Thus, in some embodiments, the memory 156 can include a random access memory (RAM), a read-only memory (ROM), a rewritable non-volatile memory, a flash memory, a hard drive, an optical disk, and the like.

[0040] Additionally, the show scene controller 110 can enable the communication circuitry 158 to interface with various other electronic devices, such as the monitoring system 111. The monitoring system 111 can use the generated 3D map and / or face direction determination for further analysis (e.g., updating the 3D map algorithm). Additionally or alternatively, the show scene controller can also enable the communication circuitry 158 to interface with components of the show scene features 160. For example, the communication circuitry 158 can enable the show scene controller 110 to communicatively couple to a network, such as a personal area network (PAN), a local area network (LAN), and / or a wide area network (WAN). Thus, in some embodiments, the show scene controller 110 can process data from the input device 152, determine the guest's face direction and position within the 3D map, determine changes (e.g., animations or effects) to be implemented, and communicate the changes to the show scene features 160 via the communication circuitry 158. For example, the processor 154 may process sensor data input from the input device 152 and then determine a control signal that enables the communications circuitry 158 to wirelessly transmit control data to the show scene feature 160 to enable actuation of the modification. In other embodiments, the communications circuitry 158 may be connected to the show scene feature 160 via a wired connection.

[0041] In some embodiments, when a control signal is sent from the show scene controller 110 to the show scene features 160, the show scene features 160 can perform a corresponding change, such as an animation performed by the animated figure 102, an image to be displayed on the display 106, or generating a visual effect using the lighting effects 104. Also, although not explicitly shown, the components used to perform the changes to the display 106, the lighting effects 104, and / or the animated figure 102 can include respective processors, respective memories, respective communication devices, and one or more actuators.

[0042] FIG. 3 is a block diagram 200 of an embodiment of the use of a face direction system 150 used to detect face direction using skeletal tracking and facial recognition and subsequently used to control an animated figure 102, according to aspects of the disclosure. The illustrated embodiment may include an animated figure 102 that changes or interacts with a guest when the guest is facing or looking at it. As illustrated, a camera 114 is positioned to observe both the ride cart 112 and guests that may come within a certain range or distance 202 (e.g., estimated by a proximity sensor 121 or a 3D map) of the animated figure 102 in the show scene that may trigger animation of the animated figure 102 when the guest is looking toward the animated figure 102. Sensory data 203 provided by the camera 114 and / or a sensor such as an illustrated RFID tag 118 may be communicated to the show scene controller 110. Based on reading the RFID by the electronic reader 123, one or more cameras 114 may be initialized or turned on to detect the direction and face direction of the ride cart 112. As mentioned above, to conserve power, presence sensors (e.g., RFID tags 118) on a series of ride carts 112 strategically placed throughout the ride 100 can be used to activate cameras 114 in sequence or at specific tracked points along the ride 100.

[0043] In this embodiment, an expected location of the ride cart 112 on the ride track 116 may be indicated by transmitting sensory data 203 to the show scene controller 110 indicating that the RFID tag 118 associated with the ride cart 112 has passed the electronic reader 123. This determination may indicate the location of the ride cart 112 on the ride track 116, such as that the ride cart 112 is slightly to the left, based on a known ride track 116 configuration. This determination may be based on multiple electronic readers 123 positioned throughout the ride track 116 such that each electronic reader 123 indicates a respective point on the ride track 116. As an example, a particular electronic reader 123 may indicate a midpoint on the ride track 116, which may include a turning ride track 116, thereby indicating that the ride cart 112 on the turning track 116 may also be turning. Additionally, the position of the ride cart 112 within the ride 100 and ride trajectory 116 may indicate a current or expected location adjacent to a show scene feature 160, such as the animated figure 102. The position of the ride cart 112 may provide context for determining a direction in which the guest may face. For example, the ride cart 112 faces forward or follows a ride direction, and thus the guest is likely facing forward. Additionally or alternatively, the position of the ride cart 112 may be determined by generating a 3D map using captured images by the camera 114 and / or the known layout of the ride 100 (e.g., the ride path 108, the ride trajectory 116, the show scene features 160, etc.).

[0044] Once the position of the ride cart 112 is determined and the camera 114 is activated, the camera 114 can focus on the view of the ride cart 112 to detect the guest's body and / or head position to determine the face direction. The show scene controller can determine the face direction using different recognition techniques such as, but not limited to, skeletal tracking and / or facial recognition as described above. As shown, there can be multiple guests in the ride cart 112 and the camera 114 can observe each guest simultaneously or individually. Thus, the show scene controller 110 can detect and track the presence of multiple guests using skeletal tracking to detect skeletal body features such as head skeleton and / or body skeleton. In some embodiments, a skeletal reading 204 sent to the show scene controller 110 can subsequently trigger face detection. The facial recognition technology can use the skeletal tracking reading to focus on the skeletal image to find the arms, chest, eyes, nose, etc. to generate the facial image 206. In this manner, the show scene controller can use the orientation of the ride cart 112 within the ride 100 in conjunction with the body / face orientation techniques as described above to determine and generate the face orientation image 208. Additionally or alternatively, the face orientation image 208 can include additional images showing predicted face rotation or position. Furthermore, the show scene controller 110 can use machine learning to update its face orientation detection determinations and predictions for guests at particular locations within the ride 100 based on recurring body and face positions and resulting face orientation determinations.

[0045] Additionally, face direction determination of one or more guests in the ride cart 112 may also be used to initialize or modify show scene features 160, such as the animated figure 102 and lighting effects 104 in the illustrated embodiment. As shown, a first guest 210 and a second guest 212 may be located closest to the animated figure 102. Based on a predefined setting (e.g., a setting defined by a designer or an operator) of the show scene controller 110, the animated figure 102 may be instructed to react to a guest who is within a certain distance 202 (e.g., 3 meters) and looking in the direction of the animated figure 102. Here, the faces of both the first guest 210 and the second guest 212 are looking towards the animated figure 102, but the first guest is within the certain distance 202 where the animated figure 102 interacts with the first guest 210, while the second guest 212 is outside the range. Specifically, a first guest 210 is within a distance 203A of 3 meters from the animated figure 102, which is within the specific distance 202 of 3 meters. Meanwhile, a second guest 212 is within a distance 203B of 26 meters from the animated figure 102, which is therefore not within the specific distance 202. Additionally, a third guest 214 is within a distance 203C of 2 meters, which is within the specific distance 202 of 3 meters from the lighting effects 104 surrounding the circus tent feature. Although the third guest 214 is within the specific distance 202, the guest's face is determined to be facing left, away from the lighting effects 104. Thus, a show scene feature can be triggered in response to a guest being determined to be within the specific distance 202 and facing toward the show scene feature. A show scene feature, such as the lighting effects 104, can conserve power using face orientation techniques described herein and continue to power off in response to no guest being determined to be facing toward the show scene feature within the specific distance 202.

[0046] Additionally, in some instances, there may be an empty ride cart (e.g., no guests in the ride cart) or a ride cart with less than the maximum number of guests. As shown in this embodiment, the ride cart 112 is followed by an empty trailing ride cart 218 as the cart continues to move forward in the ride 100. The trailing cart 218 may also include an RFID tag 118 that may be used to indicate the presence of the trailing ride cart 218. In some embodiments, when the trailing ride cart 218 is detected, the camera 114 may generate a 3D map of the ride cart 112 to further initialize skeletal tracking. However, the skeletal tracking produces a result indicating that no human skeleton is detected. Since no guests are detected, the show scene features, such as the animated figures 102 and lighting effects 104, may conserve power by remaining powered off.

[0047] 4 is a process flow diagram of a method 250 for generating a face direction determination using the show scene controller 110 of the face direction system 150 of FIG. 2 and FIG. 3. It should be noted that the method 250 may be performed as an automated procedure by a system such as the show scene controller 110 and / or the face direction system 150 including the show scene features 160. Also, while the operations described with respect to the method 250 are shown in a particular order, they may be performed in an order different from that shown and described, and such permutations of the method 250 are indeed within the scope of the present disclosure. Moreover, some steps or portions of the method 250 may be omitted and other steps may be added. The method 250 may be performed in response to a guest on a ride.

[0048] In some embodiments, the show scene controller may receive an indication that the ride cart 112 is at a location of interest (block 252). The location of interest may be a particular area in front of, behind, or near a show scene feature 160, such as the animated figure 102. Additionally or alternatively, the location may be based on ride path 108 data. The memory 156 of FIG. 2 may store ride path 108 data including information regarding the location of the ride cart 112 along the ride path 108 as a function of time and / or other factors that may be used to determine if the ride cart 112 is at a location of interest. In some embodiments, an electronic reader 123 associated with the ride cart 112 may be scanned as the ride cart 112 traverses the ride path 108. RFID signals observed along the ride path 108 may indicate the location of the ride cart 112. Additionally, in some embodiments, elapsed time and / or other factors may be used to identify the location of the ride cart 112, allowing the show scene controller 110 to determine when the ride cart 112 reaches the location of interest. In some embodiments, sensor data input (e.g., via the cameras 114, weight sensors 120, IR markers 119 and / or RFID tags 118), or the like, may be used to identify the location of the ride cart 112, and thus whether the ride cart 112 is present at the location of interest.

[0049] Additionally, the show scene controller 110 may use the sensor data input to determine the location of the ride cart 112 within the vehicle 100 (e.g., the ride path 108 or ride trajectory 116) by generating a 3D map of the location of the ride cart 112 using the techniques described above (block 254). As described above, the generation of the 3D map may be based on the location and orientation of the known ride trajectory 116, as well as known features within the vehicle 100 (e.g., the ride trajectory 116 turning toward a show scene feature at a midpoint on the ride). For example, the 3D map may generate the orientation of the ride cart 112, such as the degree of turning or banking, based on the location of the ride trajectory 116. Additionally, the guest location data within the 3D map may be based on the orientation of the ride cart 112 and the known ride cart 112 (e.g., the seats facing forward as the ride cart 112 travels along the ride path 108). Thus, the location of the ride cart 112 may also be indicative of the guest location within the vehicle 100.

[0050] After receiving an indication that the ride cart 112 is located within the area of ​​interest and determining the location of the ride cart 112, the show scene controller 110 may initialize sensors to detect one or more guests within the ride cart 112 (block 256). For example, the show scene controller 110 may output one or more signals to activate the cameras 114 used for skeletal tracking, facial recognition, and / or body recognition to begin detecting the location of the guest's face within the ride cart 112. For example, the presence of a guest may be detected by a distinguishing pattern (e.g., a continuous solid color or design) on the seat back of the ride cart 112 in the image captured by the cameras 114. The pattern may become invisible in response to a guest in the seat (e.g., a break in the continuous solid color or design due to the guest obscuring a portion of the pattern) to indicate the presence of a guest. On the other hand, if the seat is visible and the pattern is detectable, the show scene controller 110 may determine that no guest is present within the ride cart 112. This indicates that no one is present in the seat. Additionally, skeletal tracking can also be used to determine the presence of a guest such that when important joints, bones and limb intersections of the human skeleton are identified, a guest may be present within the ride cart 112. On the other hand, if skeletal joint intersections are not identified by skeletal tracking, a guest may not be present.

[0051] The show scene controller 110 may determine whether the input received from the camera 114 indicates the presence of a guest in the ride cart 112 within the observation range (decision block 258). In addition to the guest presence being determined using the camera 114 for pattern recognition or skeletal recognition on the seat, an RFID tag 118 in a guest wearable may also be scanned when the guest is seated in the ride cart 112 to indicate the guest presence.

[0052] If the show scene controller 110 receives an indication of a guest vacancy (e.g., the camera 114 captures an image of an empty cart when performing skeletal recognition), then no further face direction detection is required for at least the particular seat within the ride cart. Thus, the show scene controller 110 may stop the guest detection sensors from detecting and / or processing guests in the portion of the ride cart 112 where a vacancy is indicated (block 260).

[0053] On the other hand, if the signal indicates the presence of a guest within the seat of the ride cart 112, the show scene controller 110 may continue to perform skeletal tracking recognition or tracking to detect the guest's body and / or head (block 262). For example, skeletal tracking may generate a skeletal model taking into account the relative positions of key joints, bones and limbs of the human skeleton. As described above, joints located close to each other on the skeletal model may indicate that the body and / or head is pointing in a particular direction. Furthermore, skeletal tracking may be used in conjunction with ride cart 112 data, which may take into account show scene features at specific points on the ride trajectory 116 and along the ride path 108, to determine face orientation and make future face rotation predictions.

[0054] In addition to or instead of bone tracking used to generate a bone image of the guest, rides that use a more granular analysis of face orientation, etc., may also optionally perform face and / or body recognition to detect the guest's facial features (block 264) (as indicated by the dashed lines). As described above, face recognition may include considering facial interest points and their intersections (e.g., two eyes aligned along a horizontal line very close to the nose along a vertical line detected at or near the point between the eyes). Similarly, body detection may consider body features such as the face, arms and legs or upper body and their intersections. Both face recognition and body recognition may enable more accurate face feature and / or body determinations that may be used to determine face orientation and future rotations. As an example, a slow-paced stand-up ride where the guest stands in the ride cart 112 may utilize both body recognition and face recognition technologies.

[0055] For example, in some embodiments, a skeletal imaging algorithm can indicate the body orientation of a guest within the ride cart 112 by observing the relative proximity of skeletal joints. This information can provide an indication of the expected body orientation, and therefore face orientation. As an example, knee joints that are close to each other and pointing toward a particular side (e.g., left side) of the ride cart 112 can indicate that the guest's body and head are facing in that direction (e.g., left).

[0056] Similarly, facial recognition algorithms can indicate face orientation based on the proximity of detected facial features. For example, if a facial feature (e.g., an eye) is detected toward one side of the face, this can indicate that the face is oriented in a direction corresponding to that side of the face (e.g., eyes detected toward the left side of the face can indicate that the guest is looking left).

[0057] The show scene controller 110 can determine the guest's face direction (block 266) by using both skeletal tracking and face and / or body recognition in conjunction with the previously determined presence of the ride cart 112 and its position within the ride 100. Specifically, the ride cart direction can be used as a reference direction for the guest's body, which can be combined with the skeletal and / or face and body data to find the accurate face direction.

[0058] Additionally, in some embodiments, machine learning may be used to determine which direction the face may turn based on repeated face direction determinations made for a guest at a particular location within the ride 100 (block 267). As an example, if a guest's face direction is determined to be to the left (e.g., toward a show scene feature) at a particular point on the ride path 108, the show scene controller 110 may learn via machine learning to predict the face direction as the same direction as this location on the ride path 108. Additionally, the face direction algorithm of the show scene controller 110 may be updated using previously determined skeletal tracking, and facial and / or body recognition data, as well as machine learning on confirmed indications of the guest's actual face direction associated with the previously determined skeletal tracking, and facial and / or body recognition data. The face direction algorithm may be updated to predict or determine a more accurate face direction and if the face direction may be changing given the guest's location within the ride 100.

[0059] The face direction determination may then be used to control show scene features 160 along the ride path 108. The show scene controller 110 may determine whether the guest's face direction is within a particular distance from the show scene feature 160 (decision block 268). In other words, the show scene controller 110 may determine that the guest's face is oriented within a particular distance of the show scene feature when it is determined that the guest's face is within a particular distance of the show scene feature and that the guest's face is oriented toward the observation range of the show scene feature.

[0060] As described above, if the face direction is within a certain distance (e.g., within about 3 meters from the animated figure 102) and the user's face is directed within the viewing range of the show scene feature, the show scene feature can be activated (block 270) or updated to reflect the change. As an example, if the guest's face direction may be directed toward the animated figure 102 that is within a certain range of the guest, the show scene controller 110 can send an animation control signal to the animated figure 102. As described above, the animation control signal can be processed by a processor in the animated figure 102 to modify the animated figure 102, such as by smiling or waving using an actuator.

[0061] On the other hand, if the guest's face direction is not within a certain distance of the Shawscene feature 160 or is not directed within the viewing range of the Shawscene feature 160, the Shawscene feature will not be activated. This may provide a more personalized experience and reduce power usage and / or wear on the Shawscene feature.

[0062] In some circumstances, a guest may turn toward a show scene feature after initially detecting that the guest's direction is not within a predetermined distance of the show scene feature. Thus, further direction determinations may be performed. In such a case, the show scene controller 110 may determine whether the ride cart 112 is still in the location of interest as it travels along the ride path 108 (decision block 272). Continuing with the animated figure 102 example, if the animated figure 102 is no longer within a certain distance 202 from the guest turning toward the animated figure 102, the show scene controller 110 may cause the sensor (e.g., camera 114) to stop detecting the guest on the ride cart 112 (block 260).

[0063] On the other hand, if the ride cart 112 remains located at the point of interest, the show scene controller 110 receives an indication that the ride cart 112 is present at the point of interest (block 252) and can take subsequent steps to determine the position of the ride cart 112, initialize sensors to detect guests on the ride cart 112, etc.

[0064] As can be appreciated, state-of-the-art techniques for face orientation detection based on contextual data about the ride and / or the guest's body can provide significant benefits. For one, less granular face orientation analysis can increase the efficiency of orientation processing, which can be particularly useful in fast-paced entertainment rides. Furthermore, triggering show scene features based on detected face orientation can increase power utilization efficiency, reduce wear on show scene features, and / or provide a more personalized ride experience.

[0065] Although only certain features of the disclosure have been illustrated and described herein, many modifications and changes will occur to those skilled in the art. It is therefore to be understood that the appended claims are intended to cover all such modifications and changes that comply with the true spirit of the disclosure. It is to be understood that any of the features illustrated or described with respect to the above figures may be combined in any suitable manner.

[0066] The technology shown and claimed herein refers to and applies to tangible objects and specific examples of a practical nature that positively improve the art and are therefore not abstract, intangible, or purely theoretical. Moreover, if any claim appended at the end of this specification contains one or more elements designated as "means for [performing] ... [function]" or "step for [performing] ... [function]," such elements are to be construed pursuant to 35 U.S.C. 112(f). On the other hand, for any claim containing an element designated in any other manner, such elements are not to be construed pursuant to 35 U.S.C. 112(f).

Claims

1. 1. A face direction system for a vehicle, comprising: One or more sensors; A controller; Equipped with The controller: receiving ride cart data from the one or more sensors indicative of a presence and position of a ride cart on the vehicle; determining a vehicle cart orientation based on the vehicle cart data; receiving location data of the guest from the one or more sensors indicative of a location of the guest's body, face, or a combination thereof; determining a facial orientation, facial rotation, or a combination thereof, of the guest based at least in part on the ride cart orientation and the position data; and transmitting data indicative of the determined face orientation, face rotation, or combination thereof to a downstream controller for subsequent control based on the determined face orientation, face rotation, or combination thereof. The face direction system is configured as follows.

2. the position data includes a skeleton image determined by skeleton tracking; The face direction system according to claim 1 .

3. determining the skeletal image by skeletal tracking includes locating significant joints, bones, and limbs of the guest to generate the skeletal image; The face direction system according to claim 2 .

4. The face orientation, face rotation, or a combination thereof is predicted based on the ride cart orientation, the skeleton image, show scene features within the observation range, or a combination thereof. The face direction system according to claim 2 .

5. the controller is configured to ignore ride cart data or face position data in response to the one or more sensors detecting an absence of the guest in the ride cart. The face direction system according to claim 2 .

6. the detection of the absence of the guest in the ride cart is based on the skeletal image, pattern recognition, radio frequency identification (RFID) tags, or a combination thereof; The face direction system according to claim 5 .

7. detecting the absence of the guest in the ride cart based on a recognition pattern of a continuous solid color, a graphic, or a combination thereof on a ride cart seat; The face direction system according to claim 5 .

8. the downstream controller includes a show scene controller that controls one or more scene features based on the determined face orientation, face rotation, or a combination thereof; The show scene controller, the controller, or a combination of the show scene controller and the controller, determining whether the guest is within a specified distance of one or more show scene features of the ride; responsive to determining that the guest is within the particular distance, determining a modification of the one or more show scene features based on the face orientation; sending a control signal to trigger a change to the one or more show scene characteristics; The face direction system according to claim 1 , configured to:

9. the one or more show scene features include an animated figure, a display, a lighting effect, or a combination thereof; The face direction system according to claim 8 .

10. The modification may include an animation of the animated figure, a display image on the display, lighting of the lighting effect, or a combination thereof. The face direction system according to claim 9.

11. the change is determined in response to the guest's face direction being directed toward the one or more show scene features within a viewing range. The face direction system of claim 10.

12. receiving ride cart data from one or more sensors indicative of a presence and position of the ride cart on the ride; determining a vehicle cart orientation based on the vehicle cart data; receiving location data of the guest from the one or more sensors indicative of a location of the guest's body, face, or a combination thereof; determining a facial orientation, facial rotation, or a combination thereof, of the guest based at least in part on the ride cart orientation and position data; based on the determined face orientation, face rotation, or a combination thereof, transmitting data indicative of the determined face orientation, face rotation, or the combination thereof to a downstream controller for subsequent control; A method comprising:

13. The position data includes bone tracking images, facial recognition images, body recognition images, or a combination thereof. The method of claim 12.

14. the one or more sensors include a radio frequency identification (RFID) tag, a proximity sensor, an infrared (IR) marker, a weight sensor, a camera, or a combination thereof; The method of claim 12.

15. the RFID tags, the IR markers or a combination thereof are positioned on a ride path of the vehicle at one or more points of interest for tracking or predicting the position of the ride cart within the ride path. The method of claim 14.

16. A tangible, non-transitory, machine-readable medium containing machine-readable instructions that, when executed by one or more processors of a machine, perform the following steps: receiving ride cart data from one or more sensors indicative of a presence and position of the ride cart on the ride; determining a vehicle cart orientation based on the vehicle cart data; receiving location data of the guest from the one or more sensors indicative of a location of the guest's body, face, or a combination thereof; determining a facial rotation of the guest based at least in part on the ride cart orientation and the position data; and transmitting data indicative of the determined face orientation, face rotation, or combination thereof to a downstream controller for subsequent control based on the determined face orientation, face rotation, or combination thereof. A machine-readable medium that causes the machine to do the following.

17. The position data includes bone tracking images, facial recognition images, body recognition images, or a combination thereof.

20. The machine-readable medium of claim 16.

18. The determined facial rotation of the guest may be: a relatively less granular analysis, including facial recognition analysis, for the fast-paced portions of the ride; a relatively granular analysis using the ride cart direction, the guest's location data, or both as known cardinal directions for slower-paced portions of the ride; The machine-readable medium of claim 16,

19. a 3D map of the guest on the ride is generated based on the ride cart data, the guest position data, and the ride cart position data, the 3D map indicating the determined face direction of the guest relative to a show scene feature of the ride; 20. The machine-readable medium of claim 16.

20. The 3D map is used to accurately determine facial features, facial position, facial rotation, or a combination thereof, and the determined facial orientation is used to accurately predict facial position based on repeated facial orientation determinations made for specific portions of the vehicle within the 3D map using machine learning.

20. The machine-readable medium of claim 19.

21. a 3D map is used to determine the ride cart orientation based on known ride trajectory positions within the ride and known ride trajectory orientations; 20. The machine-readable medium of claim 16.

Images