Evaluating integrity of vehicle pose estimates via semantic tags

Abstract

The subject matter disclosed herein may relate to systems, devices, and / or methods for monitoring and / or evaluating vehicle pose integrity for use with an autonomous driving controller. A system may include a memory including one or more memory devices and a processor coupled to the one or more memory devices, the processor configured to access, from the one or more memory devices, sensor measurements generated by one or more sensors installed in a vehicle. The system may additionally observe one or more semantic categories of one or more physical entities based on the generated sensor measurements, and extract one or more semantically labeled map parameters from an electronic map, the system may additionally determine a correlation between the observed one or more semantic categories and one or more expected categories extracted from the semantically labeled map parameters.

Description

Technical Field

[0001] The subject matter disclosed herein may relate to systems, devices, and / or methods for determining the integrity of sensor signals and / or measurements for use, for example, in an autonomously operated vehicle. Background Art

[0002] In autonomous vehicle driving applications, for example, measurements from various sensors and / or measuring devices may be used in the process of estimating the vehicle's position. Based on the estimate of the vehicle's position, the autonomous vehicle driving application may be operable to reposition the vehicle or modify, for example, the vehicle's orientation, the vehicle's velocity vector, or perform some other control function or operation. By exercising such control over the vehicle's movement, the autonomous vehicle may avoid contact with obstacles (such as other vehicles, stationary or moving animals or pedestrians, road features, buildings, etc.). Failure to accurately estimate the autonomous vehicle's current position may result in potentially dangerous events, such as, for example, a wrong turn, an unsafe lane change, excessive speed, a collision, etc.

[0003] However, today's autonomous vehicle driving applications may include one or more deficiencies that, at least in some cases, could lead to such potentially dangerous events. Consequently, improving methods for localizing autonomously operated vehicles and performing autonomous driving functions appropriate to the vehicle's current location remains an active area of ​​research. Summary of the Invention

[0004] One general aspect includes a vehicle posture integrity monitor for use with an autonomous driving controller, the vehicle posture integrity monitor comprising a memory comprising one or more memory devices. The vehicle posture integrity monitor also includes a processor coupled to the one or more memory devices, the processor being configured to access sensor measurements generated by one or more sensors installed in the vehicle from the one or more memory devices. The processor is further configured to assign one or more semantic categories to one or more physical entities observed in the sensor measurements. The processor is further configured to generate expected semantic categories based on vehicle posture estimates and parameters of one or more semantic annotations from an electronic map. The processor is further configured to determine a correlation between one or more semantic categories in the observed semantic categories and one or more expected semantic categories in the expected semantic categories. Other embodiments of this aspect include a corresponding computer processor, a non-transitory computer-readable medium having instructions encoded thereon, and a method for performing the above operations.

[0005] In a particular embodiment, the one or more physical entities include a non-drivable area, a drivable area, a structure, a lane marking, a natural object or traffic sign, a dynamic object, or a combination thereof. In a particular embodiment, the processor is further configured to calculate positioning completeness based on a correlation between one or more observed semantic categories and one or more expected categories generated from the vehicle pose estimate and semantically annotated map parameters. In a particular embodiment, the processor is further configured to perform one or more operations to plan the movement of the vehicle based on the calculated positioning completeness. In a particular embodiment, the one or more operations for planning the movement of the vehicle based on the calculated positioning completeness further include one or more operations for instructing at least one autonomous driving application to perform one or more functions to affect the vehicle's motion vector. In a particular embodiment, the calculation of the correlation is based on a calculation of the number of true positives, false positives, true negatives, and false negatives for the semantic categories. In a particular embodiment, the processor is further configured to calculate one or more vehicle pose hypotheses based on a probability distribution of the vehicle pose estimate. In a particular embodiment, the processor is further configured to select one or more vehicle pose hypotheses that represent deviations from the vehicle pose estimate or deviations from the mean of the probability distribution of the vehicle pose estimate. In certain embodiments, the processor transforms each pose hypothesis into its corresponding completeness metric based on the aforementioned correlation between the observed semantic category and the expected semantic category generated from the pose hypothesis and the semantically annotated map parameters. In certain embodiments, the processor calculates the localization completeness based on a weighted average of the completeness metrics of the vehicle pose hypotheses.

[0006] Other embodiments of such aspects include corresponding computer processors, non-transitory computer-readable media having instructions encoded thereon, and methods of performing the above-described operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] The subject matter claimed is particularly pointed out and distinctly claimed in the concluding portion of the specification. However, as to organization and / or method of operation, together with objects, aspects, and / or advantages thereof, if used in conjunction with the accompanying Figure 1 The claimed subject matter may be best understood by reference to the following detailed description, read in conjunction with the accompanying drawings:

[0008] Figure 1 is a diagram depicting example sensor signal collection for assessing integrity of vehicle pose via semantic labeling according to one embodiment;

[0009] Figure 2A depicts a schematic block diagram of an example system for evaluating the completeness of a vehicle pose estimate by comparing the sensor-estimated vehicle pose with semantic labels on or derived from a digital map, according to one embodiment;

[0010] Figure 2B Describes a subprocess for computing expected semantic labels in a system for evaluating the completeness of a vehicle pose estimate, according to one embodiment;

[0011] Figures 3 and 4 depicts a flowchart of an example process for evaluating the completeness of vehicle pose estimates via semantic labels, according to an embodiment; and

[0012] Figure 5 is a schematic block diagram illustrating an example computing system environment according to one embodiment.

[0013] In the following detailed description, reference is made to the accompanying drawings, which form part of this description, wherein similar reference numerals may refer to corresponding and / or similar similar parts throughout the text. It should be understood that, such as for the simplicity and / or clarity of illustration, the drawings are not necessarily drawn to scale. For example, the dimensions of some aspects may be exaggerated relative to other aspects. In addition, it should be understood that other embodiments may be utilized. In addition, structural and / or other changes may be made without departing from the claimed subject matter. Throughout this specification, references to "claimed subject matter" refer to the subject matter intended to be covered by one or more claims or any part thereof, and are not necessarily intended to refer to a complete set of claims, a specific combination of claim sets (e.g., method claims, device claims, etc.), or a specific claim. It should also be noted that directions and / or references (e.g., such as upward, downward, top, bottom, etc.) can be used to facilitate discussion of the drawings and are not intended to limit the application of the claimed subject matter. Therefore, the following specific embodiments should not be considered to limit the claimed subject matter and / or equivalents. DETAILED DESCRIPTION

[0014] Throughout this specification, references to an implementation, an implementation, an embodiment, an embodiment, etc. mean that the specific features, structures and / or characteristics described in conjunction with the particular implementation and / or embodiment are included in at least one implementation and / or embodiment of the claimed subject matter. Thus, the appearance of such phrases, for example, in various places throughout this specification is not necessarily intended to refer to the same implementation or to any one specific implementation described. Furthermore, it should be understood that the specific aspects, structures and / or characteristics described can be combined in various ways in one or more implementations, and thus, for example, within the scope of the intended claims. Of course, in general, these and other issues vary with the context. Thus, the specific context of description and / or use provides guidance as to the inferences to be drawn.

[0015] As previously mentioned, in autonomous vehicle driving applications, for example, measurements from various sensors and / or measuring devices may be utilized in estimating the current position and orientation, or "pose," of an autonomously operating vehicle. In this context, the term "pose" refers to the current position and orientation of the vehicle. Thus, in a simple example, at a given moment, and within a predetermined error threshold, the vehicle may be oriented at a heading of 000 degrees and located at a particular set of position coordinates (e.g., GPS coordinates). However, it should be noted that the claimed subject matter is intended to encompass any method for describing the current position of the vehicle and the orientation of the vehicle, such as via other coordinate systems, with little limitation.

[0016] Autonomous vehicle driving applications can utilize estimates of the localized vehicle pose by assigning appropriate weights to output signals from onboard lidar sensors, imaging sensors (e.g., cameras), infrared sensors, radar signal sensors, ultrasonic sensors, or any other type of sensor or measurement device. Based on the estimate of the vehicle pose, the autonomous vehicle driving application can control or influence the vehicle's motion vector to avoid encountering fixed or moving obstacles that are currently in the path of the autonomously operating vehicle or estimated to be approaching the path. The weighted observations and / or measurements from the various sensors can be combined using an optimized filter of a measurement model that operates to perform cross-correlations between sensor observation and / or measurement input signals, account for sensor delays, backtrack and / or predict the estimated positions of stationary and dynamic obstacles, and eliminate or at least reduce stray sensor output signals.

[0017] Thus, based on determining that there are stationary and moving obstacles in the vehicle's intended path of travel, the autonomous vehicle driving application may apply one or more functions to control or at least influence the vehicle's speed and direction, such as applying the brakes, reducing the vehicle's speed, increasing the vehicle's speed, and / or steering the vehicle left or right to avoid encountering the obstacle. In some cases, the autonomous vehicle driving application may warn the driver that autonomous operation of the vehicle should not or cannot be performed at this time. By applying such control or influence on the vehicle's movement, an autonomously operated vehicle can safely navigate congested city streets, highways, and rural roads while avoiding contact with various types of obstacles (such as other vehicles, stationary or moving animals, natural objects, etc.). Failure to accurately estimate the current position and orientation of the autonomously operated vehicle may result in, for example, a wrong turn, an unsafe lane change, excessive speed, a collision, or may lead to many other dangerous or potentially dangerous events.

[0018] However, today's autonomous vehicle driving applications may include one or more defects that, at least in some cases, may lead to such dangerous or potentially dangerous events. Such defects may cause unacceptably high errors in the estimation of the vehicle's pose, which may be caused by obtaining and / or processing conflicting output signals from vehicle sensors and / or measurement devices installed on the autonomously operating vehicle. For example, if an on-board radar signal receiver detects the presence of an obstacle, but other on-board sensors (e.g., one or more of an on-board lidar, image capture device, ultrasonic sensor, etc.) do not detect the obstacle, the autonomous vehicle driving application may incorrectly estimate that there is no obstacle in the vehicle's driving path. In another example, the autonomous vehicle driving application may incorrectly determine that there is an obstacle when there is actually no obstacle in the vehicle's driving path. In such cases, the autonomous vehicle driving application may unnecessarily reduce the vehicle's speed and / or adjust the vehicle's driving direction to avoid the false obstacle.

[0019] To compensate for the above-mentioned shortcomings, the autonomous vehicle driving application may utilize one or more methods to determine whether a particular sensor is, for example, degraded in performance or completely failed. Based at least in part on detecting that a sensor, measuring device or other observation and / or measurement source is operating abnormally, the autonomous vehicle driving application may ignore (or at least reduce) the output signal from the suspicious sensor or measurement device. In such cases, the autonomous vehicle driving application may call a fallback or degraded operating mode, which may allow the execution of a reduced set of autonomous vehicle driving functions. However, it is understandable that, at least in some cases, the autonomous vehicle driving application may not be able to detect abnormally operating sensors, measuring devices or other observation and / or measurement sources. In such cases, the driving application may continue to access the output signals (or memory states representing such output signals) transmitted by the abnormally operating signal source, even though such signals may not accurately indicate aspects of the vehicle's positioning environment.

[0020] In certain circumstances, the inability to accurately detect abnormally functioning sensors, measuring devices, or other signal sources can present particularly problematic and / or dangerous situations. For example, an autonomous vehicle driving application might, for example, adjust the vehicle's steering in response to an onboard radar receiver correctly detecting the presence of an obstacle in the vehicle's path, but a malfunctioning lidar system is unable to accurately estimate the obstacle's location. However, in such circumstances, the autonomous vehicle driving application might steer the vehicle in the actual direction of the reported obstacle, resulting in a collision with the obstacle. Furthermore, in such circumstances, based on the failure to detect the abnormal functioning of the lidar system, such a collision could occur at an unsafe speed and without any type of warning to the driver of the autonomously operating vehicle.

[0021] Therefore, the improvement of autonomous vehicle driving application can avoid the above-mentioned scene, wherein driving application can specifically attempt to maintain the integrity of vehicle driving application via reducing or completely ignoring the output signal from specific signal source.Therefore, in specific embodiments (such as, those embodiments described herein), system-level methods can be utilized to enhance the integrity of autonomous vehicle driving application, rather than just focusing on the output signal from each abnormally running signal source.For example, in specific embodiments, autonomous vehicle driving application can utilize digital map of semantic annotation and / or the parameters derived from digital map of semantic annotation to promote system-level integrity check, rather than attempting to authenticate or verify each signal source.As used herein, the term "digital map of semantic annotation" refers to an electronic digital map, an electronic representation of at least some parameters of a digital map, or other types of digitized maps of the ground area depicting physical features in two dimensions or more.Such physical features can include drivable areas or non-drivable areas on which the features of annotation are provided. For example, a semantically annotated map can correspond to an electronic digital map indicating a driving path, such as roads (e.g., expressways, highways, toll roads, etc.), road signs (e.g., stop signs, yield signs, speed limit signs, information signs, mile markers, billboards, etc.), natural obstacles and features (e.g., trees, bodies of water, etc.), highway interchanges, exit ramps, entrance ramps, bridges, and many other features.

[0022] Thus, in certain embodiments, the pose of the autonomous vehicle may be correlated with one or more expected obstacles and / or features from a semantically annotated digital map, or expected obstacles and / or features derived from parameters of the semantically annotated digital map. Based on a relatively high correlation, or a correlation greater than a predetermined threshold, between the expected semantic category and the observed semantic category, the completeness of the localized vehicle pose may be considered relatively high. In certain embodiments, the expected semantic category is based on (or even highly dependent on) the vehicle pose estimate. Therefore, it will be appreciated that such correlation between the observed category and the semantic category can be used as a measure of the completeness of the vehicle pose estimate. Conversely, based at least in part on a relatively low correlation, or a correlation below a predetermined threshold, between the vehicle's estimated pose and the vehicle's expected position on the semantically annotated map, the completeness of the localized vehicle pose may be considered relatively low, for example. Thus, in various embodiments, based at least in part on the vehicle's calculated position completeness, an autonomous vehicle driving application may perform one or more functions to control or influence the vehicle's velocity vector. In a simple example, one or more sensors and / or measurement devices of the autonomous vehicle may estimate, for example, the presence of a "Stop" sign (or other semantically annotated traffic marker) at a particular traffic intersection. In response to estimating that the vehicle's current position strongly correlates with a particular location proximate to a "Stop" sign depicted on the semantically annotated map, it is likely that the vehicle's pose has been estimated with a relatively high degree of completeness. Thus, in short, at least in some embodiments, comparing the sensor-derived vehicle pose estimate with features expected to be present on the semantically annotated map can facilitate corroboration of the vehicle's position pose with the semantically annotated image.

[0023] In this context, the term "localization" or a variant thereof refers to the process of estimating the current position of an autonomously operated vehicle so that motion control functions for the vehicle can be planned and / or executed. Also in this context, the term "localization integrity" refers to a measure of confidence in the vehicle's pose estimate relative to the vehicle's surroundings. For example, based at least in part on a localization operation that estimates the autonomously operated vehicle to be at a particular location on a highway with relatively high localization integrity, vehicle control functions suitable for highway driving can be executed (or at least planned to be executed). Such control functions can include achieving a speed commensurate with the flow of other vehicles in traffic, maintaining a safe distance between the vehicle and vehicles ahead and behind, and so on. In another example, based at least in part on a localization operation that estimates the autonomously operated vehicle's position on a road approaching a "Stop" sign with relatively high localization integrity, an autonomous vehicle driving application can apply the vehicle's brakes to bring the vehicle to a complete stop before reaching the "Stop" sign. Conversely, based at least in part on a localization operation that estimates the autonomously operated vehicle to be at a particular location with relatively low localization integrity, the autonomous vehicle driving application can display an advisory message to the driver indicating that autonomous vehicle driving is not recommended. Alternatively or additionally, a relatively low position integrity indication may cause the autonomous vehicle driving application to enter a fallback or degraded mode of operation.

[0024] In certain embodiments, integrity checking may be able to utilize additional semantic categories beyond those used in the localization pose estimation process. In some cases, the pose estimation process may be limited to semantic categories with simple geometric parameterizations, such as lane markings, utility poles, traffic signs, and the like. In contrast, integrity checking is able to utilize a more general set of semantic categories, including dynamic objects and physical entities with more arbitrary shapes. Using such additional information and / or parameters is crucial, at least in certain embodiments, to ensure that integrity checking is not subject to the same shortcomings as the pose estimates evaluated by integrity checking.

[0025] In certain embodiments, determining the completeness of positioning by comparing the vehicle pose with semantic labels from, for example, a digital map can involve calculating a correlation coefficient that takes into account true positive rate counts, false positive rate counts, true negative rate counts, and false negative rate counts. However, as in current typical approaches to semantic annotation corresponding to categories, an imbalance in dominant semantic categories (e.g., "drivable roads," "lane markings," "traffic signs," etc.) can introduce inaccuracies in the calculation of the correlation coefficient. Therefore, certain embodiments can utilize the Matthews correlation coefficient, which advantageously takes into account positive and negative evidence (e.g., false positive counts, true positive counts, true negative counts, and false negative counts) and avoids inaccuracies introduced by dominant semantic categories. Certain embodiments can implement a computer-implemented transformation for estimating the result of applying a nonlinear transformation to a probability distribution characterized by a finite set of statistics. In one or more such embodiments, an unscented transform can be utilized to account for uncertainty in the vehicle pose estimate, which can represent a more computationally efficient approach than alternative sample-based methods (e.g., Monte Carlo methods, particle filters, histogram filters, etc.).

[0026] Certain embodiments may take into account the uncertainty or probability distribution of the pose estimate. In one or more such embodiments, pose hypotheses may be formed, where each pose hypothesis corresponds to its own correlation coefficient. A weighted average of the correlation coefficients may be calculated using a nonlinear transformation (such as an unscented transform, Monte Carlo methods, particle filters, etc.). The fused correlation coefficient may then be published as a measure of the completeness of the vehicle pose estimate.

[0027] Thus, in certain embodiments, implementing one or more of the above-described methods can achieve relatively high integrity positioning of a pose estimate for an autonomously operated vehicle. Based on determining that the vehicle pose estimate has a relatively high integrity, the autonomous vehicle driving application can determine that the observation and / or measurement equipment is being performed with a relatively high integrity. Thus, the autonomous vehicle driver can be confident that the autonomous driving application is operating within an appropriate and / or predetermined safety margin. Conversely, based on a lower correlation between obstacles observed and / or measured via sensors mounted on the vehicle and parameters extracted from the semantically annotated map, the autonomous vehicle driving application can determine that the observation and measurement equipment may be operating with a slightly lower integrity. Based on determining a slightly lower integrity of the sensor and / or measurement equipment operation, the driver of the autonomous vehicle may be less confident that the autonomous driving application is operating within an appropriate and / or predetermined safety margin.

[0028] Although specific embodiments described herein may relate to autonomously operated vehicles, the scope of the subject matter is not limited in this regard. For example, output signals from sensors and / or measuring devices can be used for any of a wide range of possible applications and / or use cases. In addition, as used herein, "sensor" and the like refer to devices and / or components that can respond to physical stimuli (such as, for example, heat, light, sound pressure, time-varying electromagnetic fields, specific motions, etc.) and / or can generate one or more signals and / or states in response to such physical stimuli. Example sensor types can include, but are not limited to, accelerometers, gyroscopes, thermometers, magnetometers, barometers, light sensors, lidar sensors, radar sensors, proximity sensors, microphones, cameras, etc., or any combination thereof.

[0029] Figure 1 is a diagram depicting example sensor signal collection for assessing vehicle pose integrity via semantic labeling according to embodiment 100. Figure 1 As shown, an autonomously operated vehicle, such as autonomously operated vehicle 102, may include a plurality of onboard sensors. Figure 1 While a particular pattern and / or a particular number of sensors are depicted and / or inferred in the embodiments of the present invention, the scope of the subject matter is not limited in this respect. For example, a system or device such as that utilized by autonomously operated vehicle 102 may include any number of sensors in any of a wide range of arrangements and / or configurations. Additionally, although Figure 1 Although depicted as a two-dimensional representation, the sensors of the autonomously operating vehicle may, for example, generate signals and / or signal groups that represent conditions in three-dimensional space surrounding the vehicle 102. In certain embodiments, the one-dimensional and / or two-dimensional sensor measurements may be combined and / or otherwise processed to produce a three-dimensional representation of the environment that at least partially surrounds or encloses the vehicle 102.

[0030] In certain embodiments, various sensors may be mounted on vehicle 102, for example, to facilitate sensing, imaging, and / or measuring different portions of the environment surrounding and / or near vehicle 102. Vehicle 102 may include multiple sensor capabilities that provide the ability to detect incoming signals, such as those corresponding to visible light, infrared energy, ultraviolet energy, RF energy, microwave and millimeter wave energy, coherent laser energy, and the like. Vehicle 102's sensors may additionally include acoustic sensors that respond to, for example, changes in sound pressure. Each sensor may utilize a different field of view to observe the environment surrounding or around vehicle 102. Example fields of view 110a, 110b, 110c, 110d, 110e, 110f, 110g, and 110h are depicted, but claimed subject matter is intended to encompass any configuration of fields of view for sensors or sensor-based measurement devices of vehicle 102.

[0031] In certain embodiments, output signals and / or output signal packets sent from onboard sensors of vehicle 102 can be utilized by at least one processor of the vehicle to facilitate semantic classification of obstacles and / or other environmental features near the vehicle. For example, such output signals can be utilized by one or more processors of vehicle 102 to autonomously guide the vehicle through the environment. Example obstacles that can be estimated to be present in the environment around vehicle 102 can include semantic categories such as other similar vehicles, as well as trucks, cyclists, pedestrians, animals, rocks, trees, lampposts, road signs, lane markings, signals, buildings, road signs, etc. Some obstacles can be classified as "static", while other objects (such as moving vehicles, trucks, cyclists, pedestrians, and animals) can be classified as "dynamic" because such obstacles move or translate through the localized environment of vehicle 102. In addition, certain features can be determined to be present in the environment around vehicle 102, which can include lane markings, traffic lights, bridges, overhead road signs, railroad tracks, etc.

[0032] In an embodiment, one or more sensors or measurement devices of the autonomously operated vehicle 102 can generate signals and / or signal packets that can represent at least a portion of the environment surrounding and / or near the vehicle 102. Additional sensors or measurement devices can generate signals and / or signal packets that represent the speed, acceleration, orientation, position (e.g., via signals from a satellite positioning system such as GPS), etc. of the vehicle 102. As described in more detail below, the sensor signals and / or signal packets can be processed to facilitate autonomous driving of the vehicle 102, such as via an autonomous vehicle driving application. In a specific implementation, for example, as the vehicle 102 advances through the environment, the sensor signals and / or signal states can be used to provide an estimate of the vehicle's pose. Based at least in part on the relatively high integrity of the vehicle's positioning parameters, the autonomous vehicle driving application can maintain control of the vehicle's velocity vector, apply brakes, or perform other functionality with a high degree of confidence that such control functions are appropriate.

[0033] Figure 2A Depicted is a schematic block diagram of an example system for evaluating the completeness of a vehicle pose estimate via comparing the sensor-estimated vehicle pose to semantic labels on or derived from a digital map, according to embodiment 200 . Figure 2A A specific non-limiting embodiment of focuses on computing a single completeness metric for the vehicle pose estimate. Figure 2A Additionally, a process for describing the probability distribution of pose estimates is described. More specifically, Figure 2AAn embodiment replaces the pose estimate with a pose hypothesis. Replacing the pose estimate with the pose hypothesis can be repeated, for example, "N" times for each of the "N" pose hypotheses. In an embodiment, the "N" integrity metrics for each of the "N" pose hypotheses can then be fused via a weighted average. The positioning integrity metric produced by the weighted average can be subsequently utilized by an autonomous vehicle driving application. It should be noted that the confusion matrix statistics refer to true positive counts, false positive counts, true negative counts, and false negative counts, which form the basis for calculating the correlation metric. The correlation metric can be used as the basis for calculating the integrity metric.

[0034] like Figure 2A As shown, vehicle 102 may include multiple signal sources represented by sensors 210. Thus, sensors 210, which may include a suite of sensors and / or measurement devices, may correspond to any number of sensors and / or devices, such as one or more of image capture devices 212, one or more of radar sensors 214, one or more of lidar sensors 216, one or more of ultrasonic sensors 218, and the like. For example, output signals from sensors 210 may include semantically annotated observations and / or measurements 215, which are used to provide input signals to computing device 220 for estimating the location environment of vehicle 102.

[0035] The computing device 220 may include or access a vehicle pose estimation module 222 to compute an estimate of the pose of the vehicle 102 to correlate with an expected position of the vehicle on or derived from the semantically annotated electronic map 225. Based at least in part on a relatively strong or high correlation between the estimated pose of the vehicle and expected features from or derived from the semantically annotated map, the position integrity of the vehicle may be considered to be relatively high. Conversely, based at least in part on a relatively weak or low correlation between the estimated pose of the vehicle and features expected to be present on the semantically annotated map, the position integrity of the vehicle may be considered to be relatively low. Figure 2AIn certain embodiments, the integrity metric measurement model 224 of the computing device 220 can be used to determine which features are likely to be present on or derived from the semantically annotated electronic map 225. Based on a relatively high correlation (or at least a predetermined threshold correlation) between the expected semantic labels 226 and the semantically annotated observations and measurements, the autonomous vehicle driving application 230 can plan and / or execute control functions for the vehicle 102. The signals and / or signal states representing the expected semantic labels 226 can then be transmitted to the confusion matrix statistics calculation module 228, which, in certain embodiments, can calculate a Matthews correlation coefficient. The Matthews correlation coefficient can utilize instances of false positives, false negatives, true positives, and true negatives, which can be utilized by the integrity metric calculation module 229. The autonomous vehicle driving application 230 can then utilize the integrity metric for the calculated vehicle pose.

[0036] Figure 2B Depicted is a sub-process for computing expected semantic labels in a system for evaluating the integrity of vehicle pose estimates according to embodiment 250. Figure 2B In certain embodiments, measurement model 256 may receive, obtain, or otherwise access signals representing parameters from semantically annotated electronic map 225 and signals representing vehicle pose estimate 254. Furthermore, measurement model 256 may be operable to perform cross-correlation between sensor observation and / or measurement input signals, account for sensor latency, backtrack and / or predict the estimated positions of stationary obstacles and features as well as dynamic obstacles, and eliminate or at least reduce spurious sensor output signals. Furthermore, measurement model 256 may backtrack and / or predict the estimated positions of stationary obstacles relative to a static reference frame, and backtrack and / or predict the estimated positions of dynamic obstacles relative to a moving reference frame. In certain embodiments, measurement model 256 may be operable to assign weights to sensor observations and / or measurements, where, for example, input signals representing a composite image from image capture sensor 212 may be weighted more heavily than individual measurements from lidar sensor 216. Measurement model 256 may include additional functionality, and claimed subject matter is not limited in this respect. Output signals from measurement model 256 may correspond to expected semantic labels 258.

[0037] In certain embodiments, an estimate of the pose of vehicle 102 may be provided in the form of a probability density function, which can be used to derive a hypothetical (e.g., most likely) pose estimate for vehicle 102. Computing device 220 may additionally observe semantic categories of static and dynamic obstacles present in the environment of vehicle 102 based on sensor measurements and / or observations. Based on the estimate of the pose of vehicle 102 or its probability density function and the sensor-derived presence of estimated static obstacles, dynamic obstacles, and / or static features, computing device 220 may analyze parameters derived from semantically annotated electronic map 225. In certain embodiments, computing device 220 may observe one or more semantic categories of sensor-estimated static obstacles, static features, and / or dynamic obstacles in the environment of vehicle 102 based on sensor observations and / or measurements. The semantic categories observed by computing device 220 may be compared to semantic categories that conform to (or correspond to) semantic labels derived from semantically annotated electronic map 225. Such sensor-estimated obstacles and / or features and their corresponding positions derived from the semantically annotated electronic map 225 may be used to confirm or refute the computed pose.

[0038] Computing device 220 may initiate a process of evaluating the integrity of the vehicle pose estimate by receiving, obtaining, or accessing observations and / or measurements 215. Based on such observations and / or measurements, computing device 220 may calculate an estimate of the pose of vehicle 102. In certain embodiments, the estimated vehicle pose may be calculated based on positioning measurements, such as signals from a satellite positioning system (e.g., GPS), an onboard compass, an onboard three-axis accelerometer, an onboard inertial navigation system, or the like. After calculating the estimated vehicle pose, computing device 220 may generate and / or observe a first set of semantic categories of static obstacles and / or static features estimated to be present in the location environment of vehicle 102 based on the sensor measurements and / or observations. Computing device 220 may then access or otherwise obtain parameters of static obstacles and / or static features that may be expected to be present in the environment of vehicle 102 from a semantically annotated electronic map 225. In an example, computing device 220 may estimate the current location of vehicle 102 to be adjacent to a particular highway interchange based on the semantically annotated observations and measurements. Based on the estimated current position of the vehicle 102, the computing device 220 may access parameters of static obstacles and / or features that may be expected to be adjacent to the estimated current position of the vehicle 102. In this example, based on the computing device 220 determining that a highway overpass exists on the semantically annotated electronic map 225, the computing device 220 may confirm with at least a predetermined confidence threshold that the pose of the vehicle 102 has been estimated within a threshold error tolerance.

[0039] In certain embodiments, vehicle 102 may utilize additional sensor measurements and / or observations (such as lidar sensors, radar sensors, etc.) to enhance the confidence of the estimated pose of vehicle 102. Accordingly, computing device 220 may utilize such additional measurements and / or observations to estimate the presence of certain additional obstacles and / or features in the vehicle. Such obstacles and / or features may include static obstacles, such as stationary vehicles, traffic lights, natural objects, etc. Additional measurements and / or observations may also enable detection of the presence of dynamic objects (such as other moving vehicles) and estimation of a more precise position of vehicle 102, such as determining that the vehicle is currently in the rightmost lane of a particular street. Computing device 220 may calculate correlations between observed semantic categories associated with obstacles and / or features based on onboard vehicle sensor measurements and / or observations and semantic labels extracted or derived from obstacles and / or features indicated on a semantically annotated digital map. A correlation module may be operable to estimate correlations between observed semantic categories and expected categories extracted from the semantically annotated digital map.

[0040] The computing device 220 can be operable to calculate any suitable correlation between the observed semantic categories and the semantic labels from the semantically annotated digital map or derived from the semantically annotated digital map. In certain embodiments, the correlation module can utilize the Matthews correlation coefficient (MCC) to determine the correlation between the observed semantic categories and the expected semantic categories. Thus, the computing device 220 can calculate the MCC according to the following expression (1):

[0041]

[0042] The MCC considers true positive (TP) counts, true negative (TN) counts, false positive (FP) counts, and false negative (FN) counts. Thus, in an embodiment, a true positive (TP) may correspond to an estimated positive match between a feature class and / or obstacle class observed by the computing device 220 in response to sensor observations and / or measurements and a feature class and / or obstacle class from the semantically annotated digital map. A true negative (TN) may correspond to a detected free space scenario where the computing device 220 has estimated that an obstacle and / or feature is not present in a portion of the vehicle environment, and the obstacle and / or feature is not present on the semantically annotated digital map. A false positive (FP) may correspond to a scenario where the computing device 220 has estimated that an obstacle class and / or feature class is present in a portion of the vehicle environment, while the semantically annotated digital map indicates that the obstacle and / or feature is not present. A false negative (FN) may correspond to a scenario where the computing device 220 has detected free space, while the semantically annotated digital map indicates that an obstacle class and / or feature class is present.

[0043] The computing device 220 may compute one or more hypotheses of the vehicle pose based on a probability distribution of the pose of the vehicle 102. In addition, the computing device 220 may compute a positive deviation or a negative deviation from the computed one or more vehicle pose hypotheses. The computing device 220 may additionally compute at least one sample point to represent a deviation from the computed one or more vehicle pose hypotheses. The computing device 220 may further compute a transformation of the distribution of the computed vehicle pose hypotheses, wherein the transformation operates to transform a nonlinear function that represents or is derived from the distribution of the one or more computed vehicle pose hypotheses. In a particular embodiment, the nonlinear function that represents or is derived from the distribution of the one or more computed vehicle pose hypotheses corresponds to the MCC of expression (1). In a particular embodiment, the transformation computed from the nonlinear function corresponds to an untraceable transformation. Methods involving untraceable transformations may be used to fuse or combine signals representing the MCC of sample points corresponding to the pose estimate of the vehicle 102. The fused or combined output signal representing the MCC (eg, representing a weighted sum of the output signals responsive to computing the MCC) can be converted to a common parameterization, such as a value between -1.0 (for complete disagreement) and +1.0 (for complete agreement).

[0044] Based on the output signal generated in response to applying the transform (such as an unscented transform), the computing device 220 can provide a most likely estimate of the vehicle pose. In addition, the computing device 220 can provide an integrity metric for the estimated pose of the vehicle 102, which represents a measure of confidence in the estimate of the vehicle pose. Based on the measure of confidence and the estimate of the vehicle pose, the autonomous vehicle driving application 230 can control and / or influence the motion vector of the vehicle 102. Such control and / or influence can include generating (or planning to generate) one or more braking commands, generating one or more acceleration commands, generating one or more left or right steering commands, or a combination thereof.

[0045] Figure 3 A flow chart depicts an example process for evaluating the integrity of vehicle pose estimates via semantic labels according to embodiment 300. In certain embodiments, Figures 3 and 4 The implementation scheme can be referred to Figure 2B305 to 330 and 405 to 420 (described below), less actions than those described in operations 305 to 330 and 405 to 420, and / or more actions than those described in operations 305 to 330 and 405 to 420. Similarly, it should be noted that the content obtained or generated by the example processes of embodiments 300 and 400 (described below) (such as, for example, input signals, output signals, operations, results, etc.) can be represented via one or more digital signals. It should also be understood that although one or more operations are shown or described simultaneously or with respect to a certain sequence, other sequences or simultaneous operations may be employed. In addition, although the following description refers to specific aspects and / or aspects shown in certain other figures, one or more operations may be performed together with other aspects and / or aspects. In an embodiment, operations 305 through 330 may be communicated as one or more signals and / or signal packets between various software, firmware, and / or hardware services executing at a computing device (e.g., such as autonomously operated vehicle 102).

[0046] The method of embodiment 300 may begin at operation 305, which includes, for example, by referring to Figure 2A Thus, operation 305 may include one or more of the sensors 210 obtaining a plurality of signal samples from a positioning sensor, such as a signal from a satellite positioning system (e.g., GPS), a compass, an accelerometer (e.g., a three-axis accelerometer), an inertial navigation system, etc. The method may continue at operation 310, which may include Figure 2A The computing device 220 obtains a pose estimate of the vehicle 102 from a localization module operating under the control of the computing device 220. The computing device 220 may generate the pose estimate by computing a weighted sum of input signals representing sensor observations and / or measurements collected at operation 305. The method may continue at operation 315, where the computing device 220 may obtain a pose estimate from the localization module. Figure 2AA first set of semantic categories of features and / or obstacles (such as static features and / or obstacles) is extracted from one or more semantically annotated electronic map parameters derived or extracted from the semantically annotated electronic map 225. The method can continue at operation 320, which can include calculating a completeness metric based on the correlation between the observed semantic labels and the expected semantic labels extracted or derived from the semantically annotated electronic map 225. The method can continue at operation 325, where, if necessary, the distribution of the distribution can be taken into account by selecting a vehicle pose hypothesis within the probability distribution of the vehicle pose estimate. The method can further include calculating a completeness metric for each pose hypothesis and then calculating a weighted average of the calculated completeness metrics. The method can continue at operation 330, which can include determining whether the pose estimate meets a threshold for use by the autonomous vehicle driving application 230 based on the completeness metric.

[0047] Figure 4 A second flow chart depicts an example process for evaluating the completeness of vehicle pose estimates via semantic labels according to embodiment 400 . Figure 4 The method may begin at operation 405, which may include obtaining or accessing sensor measurements generated by one or more sensors onboard the vehicle 102, such as from one or more memory devices. The sensor observations and / or measurements may have been semantically annotated, such as by Figure 2A The method may continue at operation 410, which may include obtaining a vehicle pose estimate from a localization module of the computing device 220. The method may continue at operation 415, which may include generating an expected semantic label using the semantically annotated electronic map and one or more vehicle pose estimates from the localization module of the computing device 220. The method may continue at operation 420, which may include calculating a correlation coefficient between the observed semantic category and the expected semantic category extracted from the semantically annotated electronic map.

[0048] Figure 5 is a diagram illustrating a computing environment according to an embodiment 500 . Figure 5 Embodiments may correspond to including autonomous vehicle driving applications such as Figure 2A Driving application 230) communicates with ( Figure 2A The computing environment of the computing device 220. Figure 5 In an embodiment of the present invention, the first device 502 and the third device 506 can facilitate the presentation of a graphical user interface (GUI) for use with a vehicle equipped with an automated vehicle driving application. In this illustration, the second device 504 can potentially provide similar functionality. Similarly, in Figure 5In an embodiment, a first device 502 may be docked with a second device 504, which may also include aspects of an onboard server computing device, for example. The processor 520 and memory 522 (which may include primary memory 525 and secondary memory 526) may communicate, for example, via a communication interface 530 and / or an input / output module 532. The term "computing device" or "computing resource" in this patent application refers to a system and / or device, such as a computing device, that includes the ability to process (e.g., perform calculations) and / or store digital content (such as electronic files, electronic documents, measurements, text, images, video, audio, etc.) in the form of signals and / or states. Therefore, in the context or environment of this patent application, a computing device may include hardware, software, firmware, or any combination thereof (in addition to the software itself). As Figure 5 The computing device 504 depicted in FIG. 5 is merely an example, and the scope of claimed subject matter is not limited to this particular example.

[0049] exist Figure 5 In the embodiment of the present invention, the first device 502 can provide one or more executable computer instruction sources in the form of, for example, physical states and / or signals (e.g., states stored in memory). The first device 502 can communicate with the second device 504, for example, via a network connection (such as via network 508). As previously mentioned, the connection, although physical, can be virtual and does not have to be tangible. Although Figure 5 The second device 504 is shown with various tangible, physical components, but the claimed subject matter is not limited to computing devices having only these tangible components, as other implementations and / or embodiments may include, for example, alternative arrangements that function differently while achieving similar results, such alternative arrangements may include additional or fewer tangible components. Rather, the examples are provided for illustration only. The scope of the claimed subject matter is not intended to be limited to the illustrative examples.

[0050] The memory 522 may include any non-transitory storage mechanism. The memory 522 may include, for example, a primary memory 525 and a secondary memory 526, and additional memory circuits, mechanisms, or combinations thereof may be used. The memory 522 may include, for example, random access memory, read-only memory, etc., such as in the form of one or more storage devices and / or systems, such as, for example, disk drives including optical drives, solid-state memory drives, etc. (to name a few examples).

[0051] Memory 522 may include one or more articles of manufacture for storing a program of executable computer instructions. For example, processor 520 may retrieve executable instructions from memory and proceed to execute the retrieved instructions. Memory 522 may also include a memory controller for accessing device-readable media 540, which may carry and / or make accessible digital content, which may include, for example, code and / or instructions that can be executed by processor 520 and / or some other device capable of executing, for example, computer instructions (such as a controller, as an example). Under the direction of processor 520, a non-transitory memory (such as a memory cell storing a physical state (e.g., a memory state)) including, for example, a program of executable computer instructions may be executed by processor 520 and may generate a signal to be communicated via, for example, a network, as previously described. The generated signal may also be stored in memory, as previously mentioned.

[0052] Memory 522 can store electronic files and / or electronic documents, such as those associated with one or more users, and can also include machine-readable media that can carry content and / or make content accessible, including, for example, code and / or instructions that can be executed by processor 520 and / or some other device capable of executing, for example, computer instructions (such as a controller, as an example). As previously mentioned, the term electronic file and / or the term electronic document are used throughout this document to refer to a set of stored memory states and / or a set of physical signals that are associated in some manner to thereby form an electronic file and / or electronic document. That is, no implicit reference is made to, for example, a specific syntax, format, and / or method used with respect to a set of associated memory states and / or a set of associated physical signals. It should also be noted that the association of memory states can be, for example, in a logical sense, and not necessarily in a tangible, physical sense. Thus, in preferred embodiments, although the signal and / or state components of an electronic file and / or electronic document are logically associated, their storage can reside, for example, in one or more different locations in tangible physical memory.

[0053] In an embodiment, Figure 5Example devices in the embodiment may include features of, for example, a client computing device and / or a remote / server computing device. It should also be noted that the term computing device, whether used as a client and / or server or otherwise, generally refers to at least a processor and memory connected by the communication bus 515. For example, "processor" is understood to refer to a specific structure, such as a central processing unit (CPU) of a computing device, which may include a control unit and an execution unit. In one aspect, a processor may include a device that interprets and executes instructions to process input signals to provide output signals. Thus, at least in the context of this patent application, computing device and / or processor are understood to refer to sufficient structure within the meaning of 35 USC § 112(f), and it is specifically intended that 35 USC § 112(f) not be implicated by the use of the terms "computing device," "processor," and / or similar terms; however, if for some reason that is not immediately apparent it is determined that the foregoing understanding does not hold, and therefore 35 USC § 112(f) is necessarily implicated by the use of the terms "computing device," "processor," and / or similar terms, then it is intended that the corresponding structure, material, and / or actions for performing one or more functions be understood and interpreted as referring to at least the functions within the meaning of 35 USC § 112(f). Figures 1 to 5 Neutralization is described in the text associated with the aforementioned figures of this patent application.

[0054] In particular implementations, obtaining measurements from sensors and / or measurement devices may include obtaining a particular set of parameters based on one or more specified parameters, which may include, for example, sorting a particular set of data elements relative to a distance from a specified point of the vehicle 102 along a specified axis and / or trajectory within a specified spatial coordinate system. Furthermore, in particular implementations, obtaining output signals from sensors and / or measurement devices may include selecting a particular set of data elements based, at least in part, on a specified range of distances from a specified point along a specified axis and / or trajectory within the specified spatial coordinate system.

[0055] Furthermore, for example, obtaining output signals from sensors and / or measurement devices based on one or more specified parameters may include grouping individual parameters of a particular set of data elements into a plurality of subsets based, at least in part, on particular individual data elements within a particular grid cell within a particular spatial coordinate system. Furthermore, in specific implementations, processing the output signals from sensors and / or measurement devices may include, for example, sorting the plurality of subsets relative to the distance of the individual grid cells from a specified point along a specified axis and / or trajectory within the particular spatial coordinate system. In specific implementations, obtaining output signals from sensors and / or measurement devices may include selecting particular data elements based, at least in part, on a specified time period, and may also include, for example, sorting the particular set of data elements relative to their temporal proximity to a specified point in time.

[0056] Unless otherwise indicated, in the context of this patent application, the term "or" (when used in an associative list such as A, B, or C) is intended to mean "A, B, and C" (used herein in an inclusive sense), as well as "A, B, or C" (used herein in an exclusive sense). With this understanding, "and" is used in an inclusive sense and is intended to mean A, B, and C; while "and / or" may be used with caution to clearly indicate that all of the foregoing meanings are intended, although such usage is not required. In addition, the terms "one or more" and / or similar terms are used to describe any feature, structure, characteristic, etc. in the singular, and "and / or" is also used to describe multiple aspects, structures, characteristics, and / or similar items and / or some other combination of aspects, structures, characteristics, and / or similar items. Likewise, the term "based on" and / or similar terms are understood to not necessarily be intended to convey an exhaustive list of factors, but rather to allow for the presence of additional factors that are not necessarily explicitly described.

[0057] In addition, for situations involving specific implementations of the claimed subject matter and subject to testing, measurement, and / or specification of degrees, it is intended that the particular situation be understood in the following manner. As an example, in a given situation, assume that the value of a physical property is to be measured. If, at least for specific implementation purposes, one of ordinary skill in the art would reasonably be likely to conceive of alternative reasonable methods of testing, measuring, and / or specification of degrees (at least with respect to properties) (continuing this example), then the claimed subject matter is intended to cover those alternative reasonable methods, unless otherwise expressly indicated. As an example, if a drawing of a measurement on an area is produced and the specific implementation of the claimed subject matter refers to taking a measurement of a slope on that area, but there are various reasonable and alternative techniques for estimating the slope on that area, then the claimed subject matter is intended to cover those reasonable alternative techniques, unless otherwise expressly indicated.

[0058] To the extent that the claimed subject matter relates to one or more specific measurements, such as with respect to physical manifestations capable of being physically measured, such as, but not limited to, temperature, pressure, voltage, current, electromagnetic radiation, etc., it is believed that the claimed subject matter does not fall within the abstract concept judicial exception to statutory subject matter. In contrast, it is asserted that physical measurements are not mental steps and, as such, are not abstract concepts.

[0059] However, it is noted that the typical measurement model employed is that one or more measurements may each comprise the sum of at least two components. Thus, for a given measurement, for example, one component may comprise a deterministic component, which in an ideal sense may comprise a physical value (e.g., found via one or more measurements) often in the form of one or more signals, signal samples, and / or states, while one component may comprise a stochastic component, which may have various sources that may be difficult to quantify. Sometimes, for example, a lack of measurement precision may affect the results of a given measurement. Thus, for the claimed subject matter, in addition to deterministic models, statistical or stochastic models may also be used as a method for identifying and / or predicting one or more measurement values ​​that may be relevant to the claimed subject matter.

[0060] For example, a relatively large number of measurements can be collected to better estimate the deterministic component. Similarly, if the measurements vary (which is often possible), it is possible that some portion of the variance can be interpreted as the deterministic component, while some portion of the variance can be interpreted as the random component. Typically, if feasible, it is desirable that the random variance associated with the measurements is relatively small. That is, typically, it may be preferable to be able to account for a reasonable portion of the measurement variation in a deterministic manner rather than in a random manner, as an aid to identification and / or predictability.

[0061] Along these lines, various techniques have been used to process one or more measurements to better estimate potential deterministic components, as well as to estimate potential random components. Of course, these techniques may vary depending on the details surrounding a given situation. However, in general, more complex problems may involve the use of more complex techniques. In this regard, as mentioned above, one or more measurements of a physical manifestation may be modeled deterministically and / or stochastically. Employing a model allows for potential identification and / or processing of the collected measurements, and / or potentially allows for estimating and / or predicting potential deterministic components, for example, with respect to subsequent measurements to be taken. A given estimate may not be a perfect estimate; however, in general, it is expected that, on average, one or more estimates may better reflect the potential deterministic components, for example, if random components that may be included in one or more of the obtained measurements are taken into account. In practical terms, of course, it is desirable to be able to generate a physically meaningful model of the process that affects the measurements to be taken, such as by an estimation method.

[0062] However, in some cases, as indicated, the potential impacts can be complex. Therefore, seeking to understand the appropriate factors to consider can be particularly challenging. Therefore, in such cases, it is not uncommon to employ heuristic approaches with respect to generating one or more estimates. Heuristics refer to the use of empirically relevant methods that can reflect the process of implementation and / or the results of implementation, such as with respect to the use of historical measurements. For example, heuristic approaches can be employed in situations where more analytical methods may be too complex and / or barely tractable. Therefore, with respect to the claimed subject matter, in example embodiments, innovative features can include heuristic approaches that can be employed to, for example, estimate and / or predict one or more measurements.

[0063] The terms "correspond," "reference," "associated," and / or similar terms refer to signals, signal samples, and / or states, such as components of a signal measurement vector, which can be stored in memory and / or used by operations to generate results that depend at least in part on the signal samples and / or signal sample states. For example, a signal sample measurement vector can be stored in a memory location and further referenced, where such reference can be embodied and / or described as a stored relationship. The stored relationship can be employed, for example, by associating (e.g., correlating) one or more memory addresses with one or more other memory addresses, and can facilitate operations that at least in part involve a combination of signal samples and / or states stored in memory for processing, such as by a processor and / or the like. Thus, in certain contexts, "associated," "referenced," and / or "corresponding" can, for example, refer to an executable process that accesses the memory contents of two or more memory locations, for example, to facilitate the performance of one or more operations between signal samples and / or states, where one or more results of the one or more operations can also be used for additional processing (such as in other operations) or can be stored in the same or other memory locations (as can be directed, for example, by executable instructions). Furthermore, the terms “get” and “read” or “store” and “write” should be understood as interchangeable terms for the corresponding operations, e.g., a result may be gotten (or read) from a memory location; similarly, a result may be stored (or written) in a memory location.

[0064] As technology has advanced, it has become more typical to employ distributed computing and / or communication methods, wherein, for example, portions of a process, such as signal processing of signal samples, may be distributed among various devices, including one or more client devices and / or one or more server devices, via a computing and / or communication network. A network may include two or more devices, such as network devices and / or computing devices, and / or devices, such as network devices and / or computing devices, may be coupled such that signal communications, such as in the form of signal packets and / or signal frames (e.g., including one or more signal samples), may be exchanged, for example, between server devices and / or client devices, as well as other types of devices, including, for example, between wired and / or wireless devices coupled via a wired and / or wireless network.

[0065] It should be understood that, for ease of description, a network device (also referred to as a networking device) may be embodied and / or described in terms of a computing device, and vice versa. However, it should also be understood that this description should in no way be construed to limit the claimed subject matter to one embodiment, such as only a computing device and / or only a network device, but rather may be embodied as a variety of devices or combinations thereof, including, for example, one or more illustrative examples.

[0066] The term "electronic file" and / or the term "electronic document" is used throughout this document to refer to a set of stored memory states and / or a set of physical signals that are associated in some manner so as to thereby at least logically form a file (e.g., an electronic file) and / or an electronic document. That is, no implicit reference is meant to be made to, for example, a specific syntax, format, and / or method used with respect to a set of associated memory states and / or a set of associated physical signals. If, for example, a specific type of file storage format and / or syntax is intended to be used, then it is explicitly referenced. It is also noted that the association of memory states may be, for example, in a logical sense, and not necessarily in a tangible, physical sense. Thus, in an embodiment, although the signal and / or state components of a file and / or electronic document will, for example, be logically associated, their storage may, for example, reside in one or more different locations in tangible physical memory.

[0067] Furthermore, in the context of this patent application, the term "parameter" (e.g., one or more parameters) refers to material that describes a collection of signal samples, such as one or more electronic documents and / or electronic files, and exists in the form of physical signals and / or physical states (e.g., memory states). For example, one or more parameters (e.g., relating to an electronic document and / or electronic file that includes an image) may include, for example, the time of day the image was captured, the latitude and longitude of the image capture device (e.g., a camera), etc. In another example, one or more parameters associated with digital content (e.g., digital content that includes technical artifacts) may include, for example, one or more authors. The claimed subject matter is intended to encompass meaningful descriptive parameters in any format, as long as the one or more parameters include physical signals and / or states. Examples of such parameters include: the name of the collection (e.g., the name of an electronic file and / or electronic document identifier), the creation technique, the purpose of the creation, the time and date of the creation, the logical path (if stored), the encoding format (e.g., the type of computer instructions, such as a markup language), and / or the standards and / or specifications used to conform to the protocol (e.g., meaning substantially conforming and / or substantially compatible) for one or more uses, etc.

[0068] Signal packet communications and / or signal frame communications (also referred to as signal packet transmissions and / or signal frame transmissions (or simply "signal packets" or "signal frames")) may be communicated between nodes of a network, where the nodes may include, for example, one or more network devices and / or one or more computing devices. By way of illustrative example, and not limitation, a node may include one or more stations employing local network addresses (e.g., in a local network address space). Similarly, devices such as network devices and / or computing devices may be associated with the node. It should also be noted that, in the context of this patent application, the term "transmission" is intended as another term for the type of signal communication that may occur in any of a variety of circumstances. Thus, it is not intended to imply a particular directionality of communication and / or a particular initiating end of a communication path used to "transmit" the communication. For example, in the context of this patent application, the mere use of the term "per se" is not intended to have a particular meaning with respect to the one or more signals being communicated, such as, for example, whether a signal is being communicated "to" a particular device, whether a signal is being communicated "from" a particular device, and / or with respect to which end of the communication path may be initiating the communication (such as, for example, whether a "push-type" signal transmission or a "pull-type" signal transmission). In the context of this patent application, a distinction is made between push-type signaling and / or pull-type signaling by which end of the communication path initiates the signaling.

[0069] For one or more embodiments, the computing device and / or the networked device may also include, for example, an autonomously operated vehicle. In addition, unless otherwise specifically stated, processes such as those described with reference to the flowcharts and / or otherwise may also be performed and / or affected in whole or in part by the computing device and / or the network device. Devices such as computing devices and / or network devices may differ in capabilities. The claimed subject matter is intended to cover a wide range of potential variations. For example, a device may include a numeric keypad and / or other limited functionality display, such as, for example, a monochrome liquid crystal display (LCD) for displaying text. However, in contrast, as another example, a network-enabled device may include a physical and / or virtual keyboard, a mass storage device, one or more accelerometers, one or more gyroscopes, a global positioning system (GPS) and / or other location identification type capabilities and / or a display with more powerful functionality, such as, for example, a touch-sensitive color 2D or 3D display.

[0070] Algorithmic descriptions and / or symbolic representations are examples of techniques used by those skilled in the art of signal processing and / or related arts to convey the substance of their work to others skilled in the art. In the context of this patent application, an algorithm is generally considered to be a self-consistent sequence of operations and / or similar signal processing leading to a desired result. In the context of this patent application, the operations and / or processing involve physical manipulations of physical quantities. Typically, although not necessarily, such quantities may take the form of electrical and / or magnetic signals and / or states capable of being stored, transferred, combined, compared, processed, and / or otherwise manipulated, such as electronic signals and / or states that constitute components of various forms of digital content (such as signal measurements, text, images, video, audio, etc.).

[0071] Primarily for reasons of commonality, it has proven convenient at times to refer to such physical signals and / or physical states as bits, values, elements, parameters, symbols, characters, terms, numbers, numerical values, measurements, contents, and the like. However, it will be understood that all of these and / or similar terms are associated with appropriate physical quantities and are merely convenient labels. Unless otherwise specifically stated, as will be apparent from the foregoing discussion, it will be understood that throughout this specification, discussions utilizing terms such as "processing," "calculating," "calculating," "determining," "establishing," "obtaining," "identifying," "selecting," "generating," and / or the like may refer to actions and / or processes of a particular apparatus (such as a special-purpose computer and / or similar special-purpose computing and / or network equipment). Thus, in the context of this specification, a special-purpose computer and / or similar special-purpose computing and / or network equipment is capable of processing, manipulating, and / or transforming signals and / or states, typically in the form of physical electronic and / or magnetic quantities, within the memory, registers, and / or other storage devices, processing devices, and / or display devices of the special-purpose computer and / or similar special-purpose computing and / or network equipment. In the context of this particular patent application, as mentioned, the term "specific apparatus" therefore includes a general-purpose computing and / or networking device (once it is programmed to perform particular functions (e.g., according to program software instructions)), such as a general-purpose computer.

[0072] In the foregoing description, various aspects of the claimed subject matter have been described. For purposes of explanation, exemplary details, such as quantities, systems, and / or configurations, have been set forth. In other cases, well-known aspects have been omitted and / or simplified so as not to obscure the claimed subject matter. Although certain aspects have been illustrated and / or described herein, numerous modifications, substitutions, variations, and / or equivalents will now occur to those skilled in the art. Therefore, it should be understood that the appended claims are intended to cover all modifications and / or variations that fall within the claimed subject matter.

Claims

1. A vehicle posture integrity monitor for use with an autonomous driving controller, the vehicle posture integrity monitor comprising: memory, said memory comprising one or more memory devices; and a processor coupled to the one or more memory devices, the processor configured to: accessing, from the one or more memory devices, sensor measurements generated by one or more sensors installed in the vehicle; observing one or more semantic categories of one or more physical entities based on the generated sensor measurements; Extracting one or more semantically annotated map parameters from an electronic map; as well as A correlation is determined between the observed one or more semantic categories and one or more expected categories extracted from the semantically annotated map parameters.

2. The vehicle posture integrity monitor according to claim 1, wherein: The one or more physical entities include: A non-drivable area, a drivable area, a structure, lane markings, a natural object or traffic sign, or a combination thereof.

3. The vehicle posture integrity monitor according to claim 1, wherein: The processor is further configured to: Position completeness is calculated based on the correlation between the observed one or more semantic categories and the one or more expected categories extracted from the semantically annotated map parameters.

4. The vehicle posture integrity monitor according to claim 3, wherein: The processor is further configured to: One or more operations are performed to plan movement of the vehicle based on the calculated positional integrity.

5. The vehicle posture integrity monitor according to claim 4, wherein: The one or more operations for planning the movement of the vehicle based on the calculated position integrity further include: One or more operations for directing at least one autonomous driving application to perform one or more functions to affect a motion vector of a vehicle.

6. The vehicle posture integrity monitor according to claim 3, wherein: The calculation of the correlation is based on the calculation of true positive rate count, false positive rate count, true negative rate count and false negative rate count.

7. The vehicle posture integrity monitor according to claim 1, wherein: The processor is further configured to: computing one or more vehicle pose hypotheses; and Compute a positive or negative deviation from one or more computed vehicle pose hypotheses.

8. The vehicle posture integrity monitor according to claim 7, wherein: The processor is further configured to: At least one sample point is calculated to represent a deviation from the calculated one or more vehicle pose hypotheses.

9. The vehicle posture integrity monitor according to claim 8, wherein: The processor is further configured to: A correlation coefficient is calculated for each of the calculated one or more vehicle pose hypotheses.

10. The vehicle posture integrity monitor according to claim 9, wherein: The calculated one or more vehicle pose hypotheses include a plurality of calculated vehicle pose hypotheses, and wherein the processor is further configured to: A transformation of the distribution of the plurality of calculated vehicle pose hypotheses is calculated.

11. The vehicle posture integrity monitor according to claim 10, wherein: The transformation operates to transform a nonlinear function that represents or is derived from a distribution of the plurality of calculated vehicle pose hypotheses.

12. The vehicle posture integrity monitor according to claim 11, wherein: The computed vehicle pose hypothesis comprises a key input for generating the one or more expected semantic categories.

13. The vehicle pose integrity monitor of claim 12, wherein the transform operates to create a weighted average of each of the one or more calculated vehicle pose hypotheses.

14. A method for providing vehicle pose integrity for use with an autonomous driving controller, the method being executed by one or more processors and comprising: accessing, from one or more memory devices, sensor measurements generated by one or more sensors installed in the vehicle; observing one or more semantic categories of one or more physical entities based on the generated sensor measurements; Extracting one or more semantically annotated map parameters from an electronic map; as well as Correlations between the observed one or more semantic categories and one or more expected categories extracted from the semantically annotated map parameters are calculated.

15. The method of claim 14, wherein the one or more physical entities comprise: A non-drivable area, a drivable area, a structure, lane markings, a natural object or traffic sign, or a combination thereof.

16. The method according to claim 14, further comprising: Position completeness is calculated based on the correlation between the observed one or more semantic categories and the one or more expected categories extracted from the semantically annotated map parameters.

17. The method according to claim 16, further comprising: One or more operations are performed to plan movement of the vehicle based on the calculated positional integrity.

18. The method according to claim 17, wherein: Performing the one or more operations to plan the movement of the vehicle further includes: One or more operations of at least one autonomous driving application are directed to perform one or more functions to affect a motion vector of the vehicle.

19. The method according to claim 16, wherein Calculating the correlation includes: Calculate the true positive rate count, calculate the false positive rate count, calculate the true negative rate count, and calculate the false negative rate count.

20. A non-transitory computer-readable medium comprising encoded program instructions for causing one or more processors of a vehicle posture integrity monitor to at least: accessing sensor measurements generated by one or more sensors installed in the vehicle; observing one or more semantic categories of one or more physical entities based on the generated sensor measurements; Extracting one or more semantically annotated map parameters from an electronic map; as well as Correlations between the observed one or more semantic categories and one or more expected categories extracted from the semantically annotated map parameters are calculated.

Images