Traffic simulator adjustment

Abstract

To provide traffic simulator adjustment.SOLUTION: The method can include determining a number of vehicles depicted in one or more images of at least a portion of a transportation network. The method may include determining, from the one or more images, respective lengths of segments in the transportation network that correspond to individual ones of the vehicles depicted in the one or more images. The method may also include determining an observed traffic density and an observed network flow corresponding to the transportation network based on the number of vehicles and the respective lengths of the segments. The method may include determining, using a traffic simulator, a traffic density estimate and a network flow estimate. The method may further include adjusting the traffic simulator based on the observed traffic density, the traffic density estimate, the observed network flow, and the network flow estimate.SELECTED DRAWING: Figure 1A

Description

[Technical Field]

[0001] The present disclosure relates generally to tuning a traffic simulator. [Background technology]

[0002] Traffic simulators may be used to help provide accurate, efficient, and / or reliable modeling of traffic flow and congestion. For example, traffic simulators can facilitate analysis of vehicle movement and interactions within a transportation network. Additionally or alternatively, traffic simulators may be deployed in a number of different contexts to help improve traffic management. For example, traffic simulators can model the impact of new infrastructure projects on existing traffic patterns. As another example, traffic simulators may be used to evaluate the impact on traffic delays and / or safety of adjusting one or more traffic control conditions (e.g., traffic signal timing, speed limits, lane closures, etc.).

[0003] The subject matter claimed in this disclosure is not limited to embodiments that solve any shortcomings or that operate only in environments such as those described above. Rather, this background is provided only to illustrate example technology areas where some embodiments described in this disclosure may be practiced. Summary of the Invention [Means for solving the problem]

[0004] According to an aspect of an embodiment, a method may include determining a number of vehicles depicted in an image of at least a portion of a transportation network. A length of each segment in the transportation network corresponding to each vehicle depicted in the image may be determined. An observed traffic density and / or an observed network flow corresponding to the transportation network may be determined based on the number of vehicles and the respective lengths of the segments. A traffic simulator may determine a traffic density estimate and / or a network flow estimate corresponding to the transportation network. The traffic simulator may be adjusted based on the observed traffic density, the traffic density estimate, the observed network flow, and / or the network flow estimate.

[0005] The object and advantages of the embodiments will be realized and attained at least by the elements, features, and combinations particularly pointed out in the claims. It is to be understood that the foregoing general description and the following detailed description are explanatory only and are not restrictive of the invention, as claimed. [Brief explanation of the drawings]

[0006] Example embodiments will be described and explained with additional specificity and detail through the use of the accompanying drawings.

[0007] [Figure 1A] 1 illustrates an exemplary embodiment of a system related to tuning a traffic simulator.

[0008] [Figure 1B] 1 illustrates an exemplary vehicle identification.

[0009] [Figure 2] 1 shows a flowchart of an exemplary method for adjusting a traffic simulator.

[0010] [Figure 3] 1 shows a block diagram of an exemplary computing system, all in accordance with one or more embodiments of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0011] Traffic simulators may be used to model transportation networks. In some situations, traffic simulators may use imagery in modeling transportation networks, which can provide a relatively large amount of traffic information corresponding to the transportation network. Because traditional traffic data sources may be relatively difficult to acquire and / or sparse (e.g., providing limited spatial coverage) due to being costly and having high maintenance demands, traffic simulators that utilize imagery (e.g., aerial and / or overhead imagery such as satellite imagery) may be more accurate and / or efficient than traffic simulators that model based on traditional traffic data sources (e.g., ground-based sensors such as loop detectors and probe vehicles). Thus, imagery-based traffic simulators may be used to improve urban planning, adjusting traffic control conditions (e.g., traffic signal timing, traffic calming measures, congestion pricing, etc.), managing emergency evacuations, etc. For example, imagery-based traffic simulators may be used to monitor and analyze the impact of natural disasters on transportation networks, plan for large-scale events, design new infrastructure projects, and / or for other applications such as providing real-time traffic management solutions.

[0012] Image-based traffic simulators may utilize artificial intelligence (AI) such as computer vision (CV). Parameters in the AI ​​model may be adjusted to improve the overall performance of the traffic simulator. In contrast, a traffic simulator that is not adjusted for improved performance may be such that the corresponding transportation network may experience undesirable consequences, such as increased traffic congestion and decreased safety. As the number of vehicles on roads increases, resulting in increased traffic congestion in the transportation network, changes to the traffic simulator may be made more frequently. An automated process for efficiently measuring such changes and generating corresponding adjustments to implement improvements to overall traffic simulator performance may help improve the traffic simulator itself and provide a more accurate traffic simulation.

[0013] Tuning a traffic simulator to improve traffic simulation performance can be error-prone due to the relatively large number of traffic conditions and / or vehicles that may be included. For example, a transportation network may include hundreds, thousands, tens of thousands, or millions of vehicles on a given day. As traffic simulators are increasingly adopted by transportation network managers, and as such traffic simulators consider more traffic conditions and / or vehicles, automated traffic simulator tuning may become increasingly desirable.

[0014] The present disclosure may relate to, among other things, methods and systems for calibrating a traffic simulator using image-based network flow estimation. Thus, traffic simulation of transportation networks may be improved by creating methods and systems for automating traffic simulator calibration according to the present disclosure. Automatically calibrating a traffic simulator based on traffic density and network flow estimation may help ensure that the traffic simulator operates at a desired level of accuracy. Additionally or alternatively, calibrating a traffic simulator using imagery may provide a more flexible and comprehensive approach to simulating traffic dynamics over wider / larger geographic areas and / or over multiple time periods.

[0015] Embodiments of the present disclosure will be described with reference to the accompanying drawings.

[0016] 1A illustrates an example embodiment of a system 100 configured to simulate a transportation network, arranged in accordance with at least some embodiments of the present disclosure. In some embodiments, the system 100 may be configured to simulate the transportation network based on an image 102.

[0017] In some embodiments, image 102 may include photographs and / or other representations generated using an optical sensor (e.g., a camera). For example, image 102 may be a collection of photographs taken by various cameras at different locations to include data for a larger area. Additionally or alternatively, image 102 may include multiple images of a single location taken over time. In some embodiments, image 102 may include one or more aerial images of an area acquired from an elevated position, such as from a position on a structure / building, and / or using an aircraft, drone, satellite, etc. For example, image 102 may include aerial imagery and / or satellite imagery. In some embodiments, image 102 may include a view of the Earth's surface (e.g., one or more roads in a transportation network). In some embodiments, image 102 may include one or more videos and / or images acquired from videos. For example, image 102 may include still photographs of a transportation network at different times captured on video.

[0018] In some embodiments, the images 102 may be relatively high resolution. For example, the images 102 may have a resolution of approximately 30 centimeters. Additionally or alternatively, the images 102 may include images with an average resolution of at least 200 pixels per inch (PPI) and / or at least 200 dots per inch (DPI). In some embodiments, the images 102 may provide broad coverage and / or detailed insight into road networks and other infrastructure and traffic patterns at a scale not achievable with traditional traffic data collection methods.

[0019] In some embodiments, the image 102 may correspond to (e.g., depict at least a portion of) a transportation network. In some embodiments, the transportation network may include interconnected routes for the movement of people and / or goods. For example, the transportation network may include one or more routes within a county, city, town, municipality, and / or geographic location of interest.

[0020] In these and other embodiments, a transportation network may include one or more links. In some embodiments, the one or more links may include segments or portions of a transportation system that can connect / enable travel between two or more points. For example, the links may include roads (e.g., highways, freeways, arterial roads, county roads, residential roads, boulevards, parkways, expressways, service roads, etc.). Additionally or alternatively, the links may be associated with trains (e.g., commuter rail lines, light rail tracks, rapid transit lines, freight rail lines, etc.), pedestrians and / or bicyclists (e.g., sidewalks, paths, crosswalks, trails, bike lanes, etc.), buses (e.g., bus lanes), water (e.g., rivers, lakes, canals, canal locks, ferry routes, ports, bridges, etc.), and / or any other transportation mode / system (e.g., aircraft runways, airports / airfields, flight paths, and / or other aviation-related links). In some embodiments, the one or more links may each include a number of lanes, a length, and / or a capacity (e.g., a maximum number of vehicles per hour operated on the link). In some embodiments, a transportation network may include one or more links without corresponding ground-based traffic measurements (e.g., traffic counts, travel times, etc. calculated at the link level). For example, a city-wide transportation network may include one or more roads where the speeds of vehicles present on the road are not regularly measured.

[0021] In some embodiments, a transportation network may be represented as a directed graph. For example, nodes of the transportation network may be intersections and / or road edges, and edges between nodes may represent links (e.g., roads) in the transportation network that form such intersections. In some embodiments, the transportation network may be configured to be traversed by one or more vehicles, including automobiles / cars, buses, trucks, motorcycles, scooters, bicycles, etc. Additionally or alternatively, the transportation network may be configured to be traversed by pedestrian traffic and / or one or more forms of public transportation (e.g., trains, streetcars, subway / metro systems, monorails, etc.).

[0022] In some embodiments, the image 102 may be acquired by the vehicle identification module 104. In some embodiments, the vehicle identification module 104 may be included in the computer vision (CV) module 112. In these and other embodiments, the vehicle identification module 104 may include code and routines configured to cause the performance of the operations described with respect to the vehicle identification module 104. Additionally or alternatively, the vehicle identification module 104 may be implemented using hardware including one or more processors, CPUs, graphics processing units (GPUs), data processing units (DPUs), parallel processing units (PPUs), microprocessors (e.g., performing or controlling the execution of one or more operations), field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), accelerators (e.g., deep learning accelerators (DLAs)), one or more programmable vision accelerators (PVAs) which may include one or more vector processing units (VPUs), one or more direct memory access (DMA) systems, one or more pixel processing engines (PPEs), etc., and / or other processor types. In these and other embodiments, vehicle identification module 104 may be implemented using a combination of hardware and software. In this disclosure, operations described as being performed by vehicle identification module 104 may include operations that vehicle identification module 104 may direct a corresponding computing system to perform. In these or other embodiments, vehicle identification module 104 may be implemented by one or more computing systems, such as those described in further detail with respect to FIG. 3.

[0023] In these and other embodiments, vehicle identification module 104 may determine the number of vehicles 106 present in one or more of images 102. For example, vehicle identification 150 in FIG. 1B may include exemplary results of determining the number of vehicles 106 depicted in each image. For example, vehicle identification 150 may analyze link segments 152 having one or more vehicles 154 to determine how many vehicles 154 are located within link segment 152 and / or where the vehicles 154 are located within link segment 152. In some embodiments, vehicle identification 150 may include creating / generating bounding shapes 156 (e.g., boxes / squares, circles / ovals, triangles, etc.) around / on top of vehicles 154. In some embodiments, bounding shapes 156 may then be counted / analyzed (e.g., via object detection and / or another CV technique) to determine the number of vehicles 154. In some embodiments, the number of vehicles 154 and the length of link segment 152 may be used to generate an observed traffic density.

[0024] Returning to FIG. 1A , in some embodiments, vehicle identification module 104 may apply a pre-trained object detector to images 102 to identify vehicle locations. In some embodiments, vehicle identification module 104 may utilize one or more artificial intelligence (AI) models to identify the number of vehicles 106 present in each image of images 102. For example, vehicle identification module 104 may use one or more CV techniques, such as image classification, object detection, semantic segmentation, instance segmentation, panoptic segmentation, visual simultaneous localization and mapping (SLAM), etc., to determine the number of vehicles 106 depicted in each image. In some embodiments, vehicle identification module 104 may generate bounding shapes around / on identified vehicles in images 102, such as bounding shape 156 of vehicle identification 150 shown in FIG. 1B . In some embodiments, the bounding shape may be geolocated / geospatially matched using geographic information embedded in the image 102 (e.g., satellite imagery generally may include embedded geographic information).

[0025] In some embodiments, the vehicle identification module 104 may classify the type of vehicle as part of determining the number of vehicles 106. For example, the vehicle identification module 104 may determine that the image contains 3,211 cars, 28 buses, and 54 trucks.

[0026] Additionally or alternatively, due to the relatively large number of pixels that may be present in one of the images 102, the vehicle identification module 104 may perform one or more object detection algorithms on a sliding window. In some embodiments, object detection by the vehicle identification module 104 (e.g., determining the number of vehicles 106 in the image) may include a sliding window of a certain size and / or other hyperparameters, such as a stride of the sliding window and a threshold for comparison with an intersection over union (IOU) metric obtained between overlapping data. For example, the vehicle identification module 104 may determine the number of vehicles 106 present in the image by performing object detection on a sliding window using an IOU threshold of about 10% of the size of the sliding window and a stride of about 50%. In some embodiments, the vehicle identification module 104 may use one or more geospatial tools, such as those for creating bounding shapes that mark detected vehicles in the image 102, such as the bounding shape 156 shown in FIG. 1B . In some embodiments, the number of vehicles 106 may be obtained by (e.g., communicated / transmitted to) the segment identification module 108.

[0027] In some embodiments, the segment identification module 108 may determine a segment length 110 within the transportation network that corresponds to an individual vehicle. In these and other embodiments, the segment identification module 108 may be included in the CV module 112. In some embodiments, the segment identification module 108 may apply one or more geospatial matching techniques and / or CV techniques to determine the segment length 110. In some embodiments, the segment identification module 108 may determine the segment length 110 based on the average length of a vehicle type. For example, if the image 102 contains 870 cars and 11 buses, and the average length of the cars is approximately 14.7 feet (4.48 meters) and the average length of the buses is approximately 35 feet (10.67 meters), the segment length 110 may be approximately 13,174 feet (4,015 meters).

[0028] In some embodiments, the CV module 112 may include code and routines configured to cause the execution of the operations described with respect to the CV module 112, the vehicle identification module 104, and / or the segment identification module 108. Additionally or alternatively, the CV module 112 may be implemented using hardware including one or more processors, CPUs, graphics processing units (GPUs), data processing units (DPUs), parallel processing units (PPUs), microprocessors (e.g., for performing or controlling the execution of one or more operations), field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), accelerators (e.g., deep learning accelerators (DLAs)), one or more programmable vision accelerators (PVAs) that may include one or more vector processing units (VPUs), one or more direct memory access (DMA) systems, one or more pixel processing engines (PPEs), etc., and / or other processor types. In these and other embodiments, the CV module 112 may be implemented using a combination of hardware and software. In this disclosure, operations described as being performed by CV module 112 may include operations that CV module 112 may direct a corresponding computing system to perform. In these or other embodiments, CV module 112 may be implemented by one or more computing systems, such as those described in further detail with respect to FIG.

[0029] In some embodiments, the number of vehicles 106 and / or the segment length 110 may be used by the CV module 112 to determine an observed traffic density 114. In some embodiments, the observed traffic density 114 may be determined for individual links and / or portions of links within a transportation network. For example, if a link (e.g., a road corresponding to a city block) may include two lanes (one in each opposing direction), each approximately 300 feet (91.44 meters) long, and as a result, eight cars are identified in an image of the link (e.g., using CV) and determined to have a segment length 110 of 128 feet (e.g., eight cars, each approximately 16 feet (4.8 meters) long), the observed traffic density 114 may be 0.2133 (e.g., 21.33% of the road may be occupied by vehicles). Additionally or alternatively, the CV module 112 may compare the segment length 110 to the total length of the links in the transportation network to determine the observed traffic density 114 for the entire transportation network. In some embodiments, observed traffic density 114 may be based on the vehicle types (e.g., as classified by segment identification module 108) of the number of vehicles 106. Additionally or alternatively, observed traffic density 114 may be based on one or more ground-based measurements of traffic.

[0030] In some embodiments, the CV module 112 may determine the observed network flow 116. For example, the CV module 112 may determine the speed and / or direction of a vehicle. For example, based on two or more images of a vehicle traveling on a link at different times, the CV module 112 may be able to determine the vehicle's speed (e.g., in a first image, a car may be identified at a first point on the road, and in a second image taken one second later, a car may be identified at a second point on the road that is one inch (2.54 centimeters) away from the first point on the image (corresponding to 88 feet (27 meters) on the road), which may be used to calculate that the vehicle is traveling at a speed of 60 miles per hour (97 kilometers per hour)). As another example, the CV module may be able to determine the direction / movement of a vehicle based on which of the images 102 the vehicle appears in. For example, a motorcycle may be identified in an image of an intersection at a first time point and then identified in an image of a road north of the intersection at a second time point, which may be used to determine that the motorcycle traveled north. In some embodiments, accuracy in determining observed network flow 116 may be improved by including one or more estimates of traffic flow / count and / or vehicle speed based on images 102 (e.g., video) collected relatively frequently (e.g., at least 24 frames per second or at least 60 frames per second) over a relatively long time window (e.g., greater than one minute). In some embodiments, the CV module 112 may communicate (e.g., transmit or send) the observed traffic density 114 and / or the observed network flow 116 determined from the images 102 to a coordination module 124 for the traffic simulator 118.

[0031] In some embodiments, system 100 may include a traffic simulator 118. Traffic simulator 118 may include code and routines configured to cause the execution of the operations described with respect to traffic simulator 118. Additionally or alternatively, traffic simulator 118 may be implemented using hardware including one or more processors, CPUs, graphics processing units (GPUs), data processing units (DPUs), parallel processing units (PPUs), microprocessors (e.g., performing or controlling the execution of one or more operations), field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), accelerators (e.g., deep learning accelerators (DLAs)), one or more programmable vision accelerators (PVAs) that may include one or more vector processing units (VPUs), one or more direct memory access (DMA) systems, one or more pixel processing engines (PPEs), etc., and / or other processor types. In these and other embodiments, traffic simulator 118 may be implemented using a combination of hardware and software. In this disclosure, operations described as being performed by traffic simulator 118 may include operations that traffic simulator 118 may direct a corresponding computing system to perform. In these or other embodiments, traffic simulator 118 may be implemented by one or more computing systems, such as those described in further detail with respect to FIG.

[0032] In some embodiments, traffic simulator 118 may be a mesoscopic simulator. For example, traffic simulator 118 may handle traffic flow dynamics for an entire transportation network as well as the choice behavior of individual travelers within the transportation network. Alternatively, traffic simulator 118 may be a microscopic simulator or a macroscopic simulator. In some embodiments, traffic simulator 118 may simulate traffic based on one or more spatiotemporal data sources. For example, traffic simulator 118 may use one or more of Global Positioning System (GPS) data, traffic sensor data, historical traffic data, incident reports, weather data, survey data, and / or any other type of spatiotemporal data to assist in simulating traffic. In some embodiments, traffic simulator 118 may generate traffic density estimates 120 and / or network flow estimates 122.

[0033] In some embodiments, traffic simulator 118 may be configured to determine traffic density estimate 120 using any suitable technique. For example, traffic simulator 118 may determine traffic density estimate 120 using one or more CV models, such as a dynamic origin-destination demand estimation (DODE) model. In some embodiments, traffic simulator 118 may be configured to determine network flow estimate 122. In these and other embodiments, network flow estimate 122 may include estimated or actual assessments of vehicle movement on a transportation network. In some embodiments, network flow estimate 122 may be generated at the link level, the route level, and / or the origin-destination level. For example, network flow estimate 122 may be an estimate of how vehicles will behave on a road during a particular period of time (e.g., simulating the same, more, or fewer vehicles on a road over time).

[0034] In some embodiments, system 100 may include coordination module 124. In some embodiments, coordination module 124 may be included in or implemented by any suitable computing system. For example, coordination module 124 may be implemented using hardware including one or more processors, central processing units (CPUs), graphics processing units (GPUs), data processing units (DPUs), parallel processing units (PPUs), microprocessors (e.g., performing or controlling the execution of one or more operations), programmable vision accelerators (PVAs) (which may include one or more direct memory access (DMA) systems and / or one or more vector or vision processing units (VPUs)), field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), accelerators (e.g., deep learning accelerators (DLAs)), and / or other processor types. Additionally or alternatively, coordination module 124 may be implemented using a combination of hardware and software. In this disclosure, operations described as being performed by coordination module 124 may include operations that coordination module 124 may instruct one or more corresponding computing systems to perform. In some embodiments, reconciliation module 124 may be configured to evaluate, compare, and / or validate traffic density estimates 120 using observed traffic densities 114. Additionally or alternatively, reconciliation module 124 may be configured to evaluate, compare, and / or validate network flow estimates 122 using observed network flows 116. In these and other embodiments, reconciliation module 124 may be implemented by one or more computing systems such as those described in further detail with respect to FIG.

[0035] In some embodiments, adjustment module 124 may determine / generate traffic simulator adjustments 126. In some embodiments, traffic simulator adjustments 126 may include instructions and / or weights that may be implemented / applied by traffic simulator 118 to modify and / or calibrate one or more traffic simulation models of traffic simulator 118.

[0036] In these and other embodiments, traffic simulator adjustment 126 may be configured to improve the performance of traffic simulator 118. For example, traffic simulator adjustment 126 may be configured to improve one or more key performance indicators (e.g., accuracy, efficiency, relevance, etc.) associated with transportation network simulation. In some embodiments, traffic simulator adjustment 126 may improve the ability of traffic simulator 118 to simulate traffic conditions based on overhead imagery (e.g., aerial and / or satellite imagery corresponding to image 102). In these and other embodiments, traffic simulator adjustment 126 may be generated using traffic density estimation 120 and network flow estimation 122. For example, in some embodiments, traffic simulator adjustment 126 may be based on reducing errors in traffic density estimation 120 and / or network flow estimation 122. In these and other embodiments, reducing errors may be based on performing one or more operations (e.g., a minimization operation, an optimization operation, etc.). For example, reducing errors in traffic density estimate 120 may include adjusting traffic simulator 118 using traffic simulator adjustment 126 so that traffic density estimate 120 matches or falls within a specified / desired range of observed traffic densities 114. As another example, reducing errors in network flow estimate 122 may include adjusting traffic simulator 118 using traffic simulator adjustment 126 so that network flow estimate 122 matches or falls within a specified / desired range of observed network flows 116.

[0037] In some embodiments, the traffic simulator adjustment 126 may be based on gradient-based optimization. For example, the adjustment module 124 may be calibrated using one or more of a traffic density estimation error, a network flow estimation error, and / or a travel time estimation error. For example, the adjustment module 124 may apply a loss function (e.g., Equation 1 below) such that the traffic simulator adjustment 126 may be generated based on the difference between the predicted / simulated traffic behavior (e.g., the traffic density estimation 120 and / or the network flow estimation 122) and the actual / desired traffic behavior (e.g., the observed traffic density 114 and the observed network flow 116). In some embodiments, the adjustment module 124 may determine the traffic simulator adjustment 126 to include one or more adjustments to match the traffic density estimation 120 with the observed traffic density 114. For example, the loss function may be used to train parameters of one or more models included in the traffic simulator used to generate the traffic simulator adjustment 126 such that the error between the predicted traffic behavior and the actual traffic behavior may be minimized. For example, the loss function may perform a minimization of the difference between modeled and observed traffic conditions by considering parameters such as link traffic counts, link travel times, and / or link densities. In some embodiments, the loss function may also be used in cases where one or more links in a transportation network do not have available ground-based measurements.

number

[0038] In some embodiments, y may be the observed count, z may be the travel time, and k may be the traffic density estimate 120. In some embodiments, L m , M m , and N mmay be an aggregation matrix used to map simulated traffic conditions to observed traffic conditions. In some embodiments, w1, w2, and w3 may be parameters that weight the relative importance of reconstructing network flow, travel time, and traffic density, respectively. In some embodiments, m may represent vehicle class / type. In some embodiments, q may be a vector of origin-destination flows. In some embodiments, p may include one or more route choice probabilities.

[0039] Additionally or alternatively, the loss function may include one or more model constraints. For example, the model constraints may be m ,ρ m ,ρ' m} m =Λ({f m} m ); f m =p m q m ;p m =Ψ m ({c m} m ,{h m} m ); and / or q m ≧0,∀m∈C. In some embodiments, f may represent a path flow. In these and other embodiments, Ψ may represent a route selection function, and Λ may represent a dynamic network loading (DNL) function. In some embodiments, the route selection function is a route selection rate (p m ) as input to determine the path (c m ) and link travel cost (h m ) may be used. In some embodiments, the Λ function is the path flow (f m ) as input, and link travel time (h m ) and / or the dynamic assignment ratio (DAR) matrix (ρ m and ρ' m) In some embodiments, Λ may be expressed as:

number

[0040] In some embodiments, traffic simulator adjustments 126 may be determined using a computation graph with one or more forward-backward algorithms. For example, a gradient-based solution algorithm may be used to solve a loss function (e.g., Dynamic Origin-Destination Demand Estimation (DODE)) on the computation graph. Thus, in some embodiments, a computing system such as computing system 300 described with respect to FIG. 3 may be used to determine (e.g., calculate or estimate) traffic simulator adjustments 126.

[0041] In some embodiments, traffic simulator adjustment 126 may be used to adjust traffic simulator 118 (e.g., one or more parameters, algorithms, techniques, and / or models of traffic simulator 118 may be adjusted). Adjusting traffic simulator 118 may be performed until a threshold and / or condition is met. For example, traffic simulator 118 may be adjusted based on traffic simulator adjustment 126 until the error in traffic density estimate 120 compared to observed traffic density 114 and / or the error in network flow estimate 122 compared to observed network flow 116 is within a specified / desired range.

[0042] Thus, in some embodiments, system 100 may be configured to adjust traffic simulator 118, thereby improving traffic simulation of a transportation network. In some embodiments, system 100 may improve the accuracy of traffic simulator 118 (e.g., by using observed traffic density 114 and observed network flow 116 obtained using CV on image 102, and using traffic density estimate 120 and network flow estimate 122 to determine traffic simulator adjustment 126, which is used to adjust traffic simulator 118). In some embodiments, system 100 may improve the ability of traffic simulator 118 to simulate traffic for a transportation network, where traffic density estimate 120 and / or network flow estimate 122 may initially be relatively poor at simulating actual / desired traffic conditions (e.g., because ground-based measurements are not available / included for one or more links). For example, traffic simulator 118 may be improved (e.g., may come closer to simulating actual / desired traffic conditions) for transportation networks that include one or more links lacking ground-based measurements because system 100 may include adjusting traffic simulator 118 based on observed traffic density 114 and / or observed network flow 116 determined using imagery 102. Thus, in these and other embodiments, system 100 may provide one or more advantages for calibrating mesoscopic traffic network models.

[0043] Modifications, additions, or omissions may be made to system 100 without departing from the scope of the present disclosure. For example, the designations of different elements in the manner described are intended to help explain the concepts described herein and are not intended to be limiting. For example, in some embodiments, image 102, vehicle identification module 104, segment identification module 108, traffic simulator 118, and / or adjustment module 124 may be depicted in the particular manner described to help explain the concepts described herein, but such depictions are not intended to be limiting. Furthermore, system 100 may include any number of other elements or may be implemented in other systems or contexts other than those described.

[0044] FIG. 2 is a flowchart of an example method 200 of adjusting a traffic simulator according to one or more embodiments of the present disclosure. Method 200 may be performed by any suitable system, apparatus, or device. For example, one or more of the operations of method 200 may be implemented by one or more devices / systems described with reference to FIG. 1A. Additionally or alternatively, one or more operations of method 200 may be performed by a computing system such as that described with reference to FIG. 3. Although shown in discrete blocks, steps and operations associated with one or more blocks of method 200 may be divided into additional blocks, combined into fewer blocks, or eliminated, depending on the specific implementation. In some embodiments, one or more non-transitory computer-readable media may be configured to store instructions that, when executed, perform method 200. In some embodiments, method 200 may be scalable to simulate / analyze traffic conditions for relatively large transportation networks (e.g., metropolitan transportation networks, regional highway systems, entire countries, etc.), which may assist in providing coordination / optimization across vast and complex infrastructures.

[0045] Method 200 may include block 202. In block 202, a number of vehicles depicted in one or more images of at least a portion of a transportation network may be determined. In some embodiments, one or more of the operations described with respect to vehicle identification module 104 of FIG. 1A may be performed as part of block 202.

[0046] Additionally or alternatively, image 102 in Figure 1A may be an example of an image. In some embodiments, the image may have a resolution of at least 40 centimeters. For example, the image may have a resolution of 30 centimeters, 20 centimeters, 10 centimeters, or higher.

[0047] In some embodiments, the images may include one or more aerial images. In these and other embodiments, the aerial images may include satellite images. In some embodiments, the satellite images may be captured from space using orbiting satellites. Additionally or alternatively, the aerial images may include one or more aerial images. In some embodiments, the aerial images may include images captured from an aerial perspective, such as from an aircraft, drone, or balloon. In some embodiments, the aerial images may include higher resolution images than the satellite images and / or other images. In some embodiments, determining the number of vehicles depicted in the images may exclude parked vehicles and / or vehicles not located within the transportation network (e.g., vehicles identified as being in a parking lot, garage, driveway, garage, roadside, etc.). For example, vehicles more than three lanes away from the centerline and / or more than 118 feet (36 meters) away from the centerline in links identified as being within the transportation network may be excluded.

[0048] In block 204, the length of each of the segments in the transportation network corresponding to each of the vehicles depicted in the image may be determined from the image. In some embodiments, the length of each of the segments may be determined based at least in part on the vehicle type of each vehicle determined to be depicted in the image. In these and other embodiments, one or more of the operations described with respect to segment identification module 108 of FIG. 1A may be performed as part of identifying the length of the segments.

[0049] At block 206, an observed traffic density and an observed network flow corresponding to the transportation network may be determined based on the number of vehicles and the respective lengths of the segments. In some embodiments, determining the observed traffic density and the observed network flow corresponding to the transportation network may include determining the observed traffic density and / or the observed network flow of one or more links in the transportation network without corresponding ground-based traffic measurements. In some embodiments, determining the observed traffic density and / or the observed network flow corresponding to the transportation network may include using one or more computer vision (CV) techniques. For example, determining the observed traffic density and / or the observed network flow may include image classification, object detection, semantic segmentation, instance segmentation, panoptic segmentation, and / or visual simultaneous localization and mapping (SLAM). For example, determining the observed traffic density and / or the observed network flow may include using object detection performed over a sliding window of a certain size. In some embodiments, determining the observed traffic density may be based on each classified vehicle type. In these and other embodiments, one or more of the operations described with respect to the CV module 112 of FIG. 1A may be performed as part of determining observed traffic density and / or observed network flow.

[0050] At block 208, a traffic simulator may be used to determine a traffic density estimate and / or a network flow estimate corresponding to the transportation network. In some embodiments, the traffic density estimate and / or the network flow estimate may be determined using the network flow estimation error, the travel time estimation error, and / or the traffic density estimation error. In these and other embodiments, one or more of the operations described with respect to the traffic simulator 118 of FIG. 1A may be performed as part of determining the traffic density estimate and / or the network flow estimate.

[0051] At block 210, the traffic simulator may be adjusted based on the observed traffic density, traffic density estimates, observed network flows, and / or network flow estimates. In some embodiments, adjusting the traffic simulator based on the observed traffic density, traffic density estimates, observed network flows, and / or network flow estimates may improve the calibration accuracy of the traffic simulator. For example, links that lack ground-based measurements of traffic flow and / or travel times may benefit from the improved calibration accuracy obtained by adjusting the traffic simulator using the observed traffic density, traffic density estimates, observed network flows, and network flow estimates. In these and other embodiments, one or more of the operations described with respect to adjustment module 124 of FIG. 1A may be performed as part of adjusting the traffic simulator.

[0052] Modifications, additions, or omissions may be made to method 200 without departing from the scope of the present disclosure. For example, the designations of various elements in the described manner are intended to help explain the concepts described herein and are not limiting. Furthermore, method 200 may include any number of other elements or may be implemented in other systems or contexts other than those described.

[0053] 3 illustrates an exemplary computing system 300 in accordance with one or more embodiments of the present disclosure. The computing system 300 may include a processor 302, a memory 304, a data storage 306, and / or a communication unit 308, all or one or more of which may be communicatively coupled. For example, the vehicle identification module 104, the segment identification module 108, the CV module 112, the traffic simulator 118, and / or the calibration module 124 of FIG. 1A may be implemented as or configured to utilize a computing system consistent with the computing system 300.

[0054] In general, processor 302 may include any suitable special-purpose or general-purpose computer, computing entity, or processing device, including various computer hardware or software modules, and may be configured to execute instructions stored on any applicable computer-readable storage medium. For example, processor 302 may include a microprocessor, microcontroller, digital signal processor (DSP), application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or any other digital or analog circuitry configured to interpret and / or execute program instructions and / or process data.

[0055] 3 as a single processor, it is understood that processor 302 may include any number of processors distributed across any number of networks or physical locations configured to individually or collectively perform any number of operations described in this disclosure. In some embodiments, processor 302 may interpret and / or execute program instructions and / or process data stored in memory 304, data storage 306, or memory 304 and data storage 306. In some embodiments, processor 302 may fetch program instructions from data storage 306 and load program instructions into memory 304.

[0056] After the program instructions are loaded into memory 304, processor 302 may execute the program instructions, such as instructions that cause computing system 300 to perform some of the operations of method 200 of Figure 2. For example, computing system 300 may execute the program instructions to determine a number of vehicles depicted in one or more images of at least a portion of a transportation network, determine from the images the lengths of each of the segments in the transportation network corresponding to each of the vehicles depicted in the images, determine an observed traffic density and an observed network flow corresponding to the transportation network based on the number of vehicles and the lengths of each of the segments, determine a traffic density estimate and a network flow estimate using a traffic simulator; and / or adjust the traffic simulator based on the observed traffic density, traffic density estimate, observed network flow, and / or network flow estimate of method 200 of Figure 2.

[0057] Memory 304 and data storage 306 may include one or more computer-readable storage media for storing computer-executable instructions or data structures. Such computer-readable storage media may be any available media that can be accessed by a general-purpose or special-purpose computer, such as processor 302. In some embodiments, computing system 300 may or may not include either memory 304 or data storage 306.

[0058] By way of example, and not limitation, such computer-readable storage media may include non-transitory computer-readable storage media including random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disk storage, magnetic disk storage or other magnetic storage devices, flash memory devices (e.g., solid-state memory devices), or any other storage medium that may be used to store desired program code in the form of computer-executable instructions or data structures and that can be accessed by a general-purpose or special-purpose computer. Combinations of the above may also be included within the scope of computer-readable storage media. Computer-executable instructions may include, for example, instructions and data configured to cause processor 302 to perform a particular operation or group of operations.

[0059] The communications unit 308 may include any component, device, system, or combination thereof configured to transmit or receive information over a network. In some embodiments, the communications unit 308 may communicate with other devices at other locations, the same location, or other components within the same system. For example, the communications unit 308 may include a modem, a network card (wireless or wired), an optical communications device, an infrared communications device, a wireless communications device (such as an antenna), and / or a chipset (such as a Bluetooth device, an 802.6 device (e.g., a metropolitan area network (MAN)), a WiFi device, a WiMax device, a cellular communications facility, etc.). The communications unit 308 may enable data to be exchanged with a network and / or any other device or system described in this disclosure. For example, the communications unit 308 may allow the computing system 300 to communicate with other systems, such as computing devices and / or other networks.

[0060] Those skilled in the art, after reviewing the present disclosure, will recognize that modifications, additions, or omissions may be made to computing system 300 without departing from the scope of the present disclosure. For example, computing system 300 may include more or fewer components than explicitly shown and described.

[0061] The foregoing disclosure is not intended to limit the disclosure to the precise form or particular field of use disclosed. Accordingly, various alternative embodiments and / or modifications to the present disclosure, whether expressly described or implied herein, are contemplated as possible in light of the present disclosure. While embodiments of the present disclosure have been thus described, it will be recognized that changes can be made in form and detail without departing from the scope of the present disclosure. Accordingly, the present disclosure is limited only by the claims.

[0062] In some embodiments, the different components, modules, engines, and services described herein may be implemented as objects or processes (e.g., as separate threads) executing on a computing system. While some of the systems and methods described herein are generally described as being implemented in software (stored and / or executed on general-purpose hardware), specific hardware implementations or combinations of software and specific hardware implementations are also possible and contemplated.

[0063] According to common practice, various features illustrated in the drawings may not be drawn to scale. The illustrations presented in this disclosure are not intended to be actual diagrams of any particular apparatus (e.g., device, system, etc.) or method, but merely idealized representations used to describe various embodiments of the present disclosure. Thus, dimensions of various features may be arbitrarily increased or decreased for clarity. In addition, some of the drawings may be simplified for clarity. Thus, the drawings may not show all of the components of a given apparatus (e.g., device) or all operations of a particular method.

[0064] The terms used in this specification, and particularly in the appended claims (e.g., the body of the appended claims), are generally intended as "open" terms (e.g., the term "including" should be interpreted as "including, but not limited to," the term "having" should be interpreted as "having at least...," the term "including" should be interpreted as "including, but not limited to," etc.).

[0065] Furthermore, where a specific number of introduced claim recitations is intended, such intention will be expressly recited in the claim; absent such recitation, no such intention exists. For example, to aid in understanding, the following appended claims may include the use of the introductory phrases "at least one" and "one or more" to introduce claim recitations. However, the use of such phrases should not be construed as implying that introducing a claim recitation with the indefinite article "a" or "an" limits any particular claim containing such an introduced claim recitation to embodiments containing only one such recitation. This is true even when the same claim includes the introductory phrase "one or more" or "at least one" and an indefinite article such as "a" or "an." (For example, "a" and / or "an" should be construed to mean "at least one" or "one or more.") The same applies to the use of definite articles used to introduce claim recitations.

[0066] Furthermore, even when a particular number of introduced claim recitations is explicitly recited, it is understood that such recitation should be interpreted to mean at least the recited number (e.g., the recitation "two recitations" without other modifiers means at least two recitations, or more than two recitations). Furthermore, when a convention similar to "at least one of A, B, and C, etc." or "one or more of A, B, and C, etc." is used, such construction is generally intended to include A only, B only, C only, A and B together, A and C together, B and C together, or A, B, and C together, etc. For example, use of the term "and / or" is intended to be interpreted in this manner.

[0067] Furthermore, any disjunctive phrase presenting two or more alternative terms, whether in the present document, claims, or drawings, should be understood to contemplate the inclusion of one of those terms, either of those terms, or both terms. For example, the phrase "A or B" should be understood to include the possibilities of "A" or "B" or "A and B."

[0068] Furthermore, the use of terms such as “first,” “second,” and “third” is not used herein to necessarily imply a particular order or number of elements. In general, terms such as “first,” “second,” and “third” are used as general identifiers to distinguish between different elements. Unless otherwise indicated, terms such as “first,” “second,” and “third” should not be understood to imply a particular order. Furthermore, unless otherwise indicated, terms such as “first,” “second,” and “third” should not be understood to imply a particular number of elements. For example, a first widget may be described as having a first side, and a second widget may be described as having a second side. Use of the term “second side” with respect to a second widget is intended to distinguish such side of the second widget from the “first side” of the first widget and may not imply that the second widget has two sides.

[0069] All examples and conditional language set forth herein are intended for educational purposes to help the reader understand the invention and the concepts contributed by the inventor to further the art, and should be construed without limitation to such specifically set forth examples and conditions. Although embodiments of the present disclosure have been described in detail, it should be understood that various changes, substitutions, and alterations can be made to the present disclosure without departing from the spirit and scope of the present disclosure.

[0070] The following additional notes are provided regarding the embodiments including the above examples. (Appendix 1) determining a number of vehicles depicted in one or more images of at least a portion of the transportation network; determining a length of each of the segments in the transportation network corresponding to each of the vehicles depicted in the one or more images; determining an observed traffic density and an observed network flow corresponding to the transportation network based on the number of vehicles and the length of each of the segments; determining, using a traffic simulator, a traffic density estimate and a network flow estimate corresponding to the transportation network; calibrating the traffic simulator based on the observed traffic density, the traffic density estimate, the observed network flow, and the network flow estimate; Including, method. (Appendix 2) 2. The method of claim 1, wherein the one or more images have a resolution of at least 30 centimeters and include one or more of aerial or satellite images. (Appendix 3) 2. The method of claim 1, wherein the transportation network includes one or more links without corresponding ground-based traffic measurements. (Appendix 4) Determining observed traffic densities and observed network flows corresponding to said transportation network comprises: Image classification, object detection, Semantic segmentation, Instance segmentation. Panoptic segmentation, or Visual simultaneous localization and mapping

[0023] using one or more computer vision techniques, including one or more of: The method described in Appendix 1. (Appendix 5) 2. The method of claim 1, wherein determining observed traffic densities and observed network flows corresponding to the transportation network includes using object detection performed over a sliding window of a certain size. (Appendix 6) 2. The method of claim 1, further comprising classifying each of the vehicles by type, and determining an observed traffic density and an observed network flow corresponding to the transportation network based on the classified vehicle type. (Appendix 7) Adjusting the traffic simulator includes: determining one or more of a network flow estimation error, a travel time estimation error, or a traffic density estimation error; The method described in Appendix 1. (Appendix 8) 1. A system having one or more processors configured to perform operations, the operations including: The observed traffic density and observed network flows corresponding to the transportation network: the number of vehicles depicted in the image or images; the type of each of the vehicles depicted in the one or more images; or a length of each of the segments in the transportation network corresponding to the vehicle, determined based on the one or more images; Determined based on one or more of the following: adjusting one or more of the traffic density estimates or the network flow estimates determined by the traffic simulator based on the observed traffic density and the observed network flow; system. (Appendix 9) 9. The system of claim 8, wherein the one or more images have a resolution of at least 30 centimeters and include one or more of aerial or satellite images. (Appendix 10) 9. The system of claim 8, wherein the transportation network includes one or more links without corresponding ground-based traffic measurements. (Appendix 11) Determining said observed traffic density and observed network flow comprises: Image classification, object detection, Semantic segmentation, Instance segmentation. Panoptic segmentation, or Visual simultaneous localization and mapping

[0023] using one or more computer vision techniques, including one or more of: 10. The system of claim 8. (Appendix 12) 9. The system of claim 8, wherein determining the observed traffic density and the observed network flow includes using object detection performed over a sliding window of a certain size. (Appendix 13) 9. The system of claim 8, wherein the operations further include classifying each of the vehicles by type, and wherein determining the observed traffic density and observed network flow is based on the classified vehicle type. (Appendix 14) adjusting one or more of the traffic density estimates or network flow estimates determined by said traffic simulator; determining one or more of a network flow estimation error, a travel time estimation error, or a traffic density estimation error; 10. The system of claim 8. (Appendix 15) One or more non-transitory computer-readable storage media configured to store instructions that, in response to being executed, cause a system to perform operations, the operations including: determining a number of vehicles depicted in the image of at least a portion of the transportation network; determining the length of each of the segments in the transportation network corresponding to each of the vehicles depicted in the image; determining an observed traffic density and an observed network flow corresponding to the transportation network based on the number of vehicles and the length of each of the segments; determining, using a traffic simulator, a traffic density estimate and a network flow estimate corresponding to the transportation network; calibrating the traffic simulator based on the observed traffic density, the traffic density estimate, the observed network flow, and the network flow estimate; Including, One or more non-transitory computer-readable storage media. (Appendix 16) 16. The one or more non-transitory computer-readable storage media of claim 15, wherein the images have a resolution of at least 30 centimeters and include one or more of aerial imagery or satellite imagery. (Appendix 17) 16. The one or more non-transitory computer-readable storage media of claim 15, wherein the transportation network includes one or more links without corresponding ground-based traffic measurements. (Appendix 18) Determining said observed traffic density and observed network flow comprises: Image classification, object detection, Semantic segmentation, Instance segmentation. Panoptic segmentation, or Visual simultaneous localization and mapping

[0023] using one or more computer vision techniques, including one or more of: 16. One or more non-transitory computer-readable storage media as described in Clause 15. (Appendix 19) 16. The one or more non-transitory computer-readable storage media of claim 15, wherein the operations further include classifying each of the vehicles by type, and wherein determining the observed traffic density and observed network flow is based on the classified vehicle type. (Appendix 20) Adjusting the traffic simulator includes: determining one or more of a network flow estimation error, a travel time estimation error, or a traffic density estimation error; 16. One or more non-transitory computer-readable storage media as described in Clause 15. [Explanation of symbols]

[0071] 102 images 104 Vehicle Identification Module 106 vehicles 108 Segment Identification Module 110 segment length 112 Computer Vision (CV) Module 114 Observed traffic density 116 Observed Network Flows 118 Traffic Simulator 120 Traffic density estimation 122 Network Flow Estimation 124 Adjustment Module 126 Traffic Simulator Adjustment 202 determining the number of vehicles depicted in one or more images of at least a portion of the transportation network; 204. Determine, from the one or more images, the length of each of the segments in the transportation network corresponding to each of the vehicles depicted in the one or more images. Determine the observed traffic density and observed network flow corresponding to the transportation network based on the number of 206 vehicles and the length of each of the segments. 208 Use a traffic simulator to determine traffic density estimates and network flow estimates for the transportation network. 210 Calibrate the traffic simulator based on observed traffic density, traffic density estimates, observed network flows, and network flow estimates 300 Computing Systems 302 processor 304 memory 306 Data Storage 308 Communication Unit

Claims

1. determining a number of vehicles depicted in one or more images of at least a portion of the transportation network; determining the length of each of the segments in the transportation network corresponding to each of the vehicles depicted in the one or more images; determining an observed traffic density and an observed network flow corresponding to the transportation network based on the number of vehicles and the length of each of the segments; determining traffic density estimates and network flow estimates corresponding to the transportation network using a traffic simulator; adjusting the traffic simulator based on the observed traffic density, the traffic density estimate, the observed network flow, and the network flow estimate; Including, method.

2. The method of claim 1 , wherein the one or more images are at least 30 centimeter resolution and include one or more of aerial or satellite imagery.

3. The method of claim 1 , wherein the transportation network includes one or more links without corresponding ground-based traffic measurements.

4. Determining an observed traffic density and an observed network flow corresponding to the transportation network includes: Image classification, object detection, Semantic segmentation, Instance segmentation. Panoptic segmentation, or Visual simultaneous localization and mapping [0023] using one or more computer vision techniques, including one or more of: The method of claim 1.

5. The method of claim 1 , wherein determining observed traffic density and observed network flow corresponding to the transportation network includes using object detection performed over a sliding window of a certain size.

6. 10. The method of claim 1, further comprising classifying each of the vehicles by type, and determining an observed traffic density and an observed network flow corresponding to the transportation network based on the classified vehicle type.

7. Adjusting the traffic simulator includes: determining one or more of a network flow estimation error, a travel time estimation error, or a traffic density estimation error; The method of claim 1.

8. 1. A system having one or more processors configured to perform operations, the operations including: The observed traffic density and observed network flows corresponding to the transportation network are: the number of vehicles depicted in the image or images; the type of each of the vehicles depicted in the one or more images; or a length of each of the segments in the transportation network corresponding to the vehicle, determined based on the one or more images; determining based on one or more of the following: adjusting one or more of a traffic density estimate or a network flow estimate determined by a traffic simulator based on the observed traffic density and the observed network flow. system.

9. The system of claim 8 , wherein the one or more images are at least 30 centimeter resolution and include one or more of aerial or satellite imagery.

10. The system of claim 8 , wherein the transportation network includes one or more links without corresponding ground-based traffic measurements.

11. Determining the observed traffic density and the observed network flow includes: Image classification, object detection, Semantic segmentation, Instance segmentation. Panoptic segmentation, or Visual simultaneous localization and mapping [0023] using one or more computer vision techniques, including one or more of: The system of claim 8.

12. The system of claim 8 , wherein determining the observed traffic density and the observed network flow includes using object detection performed over a sliding window of a certain size.

13. 9. The system of claim 8, wherein the operations further include classifying each of the vehicles by type, and wherein determining the observed traffic density and observed network flow is based on the classified vehicle type.

14. Adjusting one or more of the traffic density estimates or the network flow estimates determined by the traffic simulator includes: determining one or more of a network flow estimation error, a travel time estimation error, or a traffic density estimation error; The system of claim 8.

15. One or more non-transitory computer-readable storage media configured to store instructions that, in response to being executed, cause a system to perform operations, the operations including: determining a number of vehicles depicted in the image of at least a portion of the transportation network; determining the length of each of the segments in the transportation network corresponding to each of the vehicles depicted in the image; determining an observed traffic density and an observed network flow corresponding to the transportation network based on the number of vehicles and the length of each of the segments; determining traffic density estimates and network flow estimates corresponding to the transportation network using a traffic simulator; adjusting the traffic simulator based on the observed traffic density, the traffic density estimate, the observed network flow, and the network flow estimate; Including, One or more non-transitory computer-readable storage media.

16. 16. The one or more non-transitory computer-readable storage media of claim 15, wherein the images are at least 30 centimeter resolution and include one or more of aerial or satellite images.

17. 16. The one or more non-transitory computer-readable storage media of claim 15, wherein the transportation network includes one or more links without corresponding ground-based traffic measurements.

18. Determining the observed traffic density and the observed network flow includes: Image classification, object detection, Semantic segmentation, Instance segmentation. Panoptic segmentation, or Visual simultaneous localization and mapping [0023] using one or more computer vision techniques, including one or more of:

16. One or more non-transitory computer-readable storage media according to claim 15.

19. 16. The one or more non-transitory computer-readable storage media of claim 15, wherein the operations further include classifying each of the vehicles by type, and wherein determining the observed traffic density and observed network flow is based on the classified vehicle type.

20. Adjusting the traffic simulator includes: determining one or more of a network flow estimation error, a travel time estimation error, or a traffic density estimation error; 16. One or more non-transitory computer-readable storage media according to claim 15.

Images