Traffic Light Detection and Classification for Autonomous Vehicles

Abstract

Perceive the driving environment based on sensor data obtained from multiple sensors installed on the ADV, including detecting traffic lights, wherein the multiple sensors include at least one image sensor. Apply a first sensor setting (402) to the at least one image sensor to capture a first frame, and apply a second sensor setting (403) to the at least one image sensor to capture a second frame. Determine the color of the traffic light based on the sensor data of the at least one image sensor in the first frame. Control the ADV for autonomous driving according to the color of the traffic light determined based on the sensor data of the at least one image sensor in the first frame and the driving environment perceived based on the sensor data of the at least one image sensor in the second frame.

Description

Technical Field

[0001] Embodiments of the present disclosure generally relate to operating an autonomous driving vehicle (ADV). More specifically, embodiments of the present disclosure relate to traffic light detection and classification for an ADV. Background Art

[0002] A vehicle operating in an autonomous driving mode (e.g., driverless) can free the occupants, especially the driver, from some driving-related responsibilities. When operating in an autonomous driving mode, the vehicle can use on-board sensors to navigate to various locations, allowing the vehicle to travel with minimal human interaction or in some cases without any passengers.

[0003] Motion planning and control are key operations in autonomous driving. Traffic light detection and classification are important for motion planning and control of an ADV. However, it is difficult to identify a red traffic light due to color artifacts, for example, in a dark / cloudy environment. Many cases of failure to identify a red traffic light have been reported. Summary of the Invention

[0004] Embodiments of the present disclosure provide computer-implemented methods, non-transitory machine-readable media, data processing systems, and computer program products for operating an autonomous driving vehicle (ADV).

[0005] In a first aspect, some embodiments of the present disclosure provide a computer-implemented method for operating an autonomous driving vehicle (ADV). The method includes: perceiving a driving environment based on sensor data obtained from a plurality of sensors mounted on the ADV, including detecting a traffic light, the plurality of sensors including at least one image sensor; applying a first sensor setting to the at least one image sensor to capture a first frame; applying a second sensor setting to the at least one image sensor to capture a second frame; determining a color of the traffic light based on sensor data of the at least one image sensor in the first frame; and controlling the ADV to drive autonomously according to the color of the traffic light determined based on sensor data of the at least one image sensor in the first frame and the driving environment perceived based on sensor data of the at least one image sensor in the second frame.

[0006] In a second aspect, some embodiments of the present disclosure provide a non-transitory machine-readable medium storing instructions that, when executed by a processor, cause the processor to perform operations, the operations including: perceiving a driving environment based on sensor data obtained from a plurality of sensors installed on an ADV, including detecting traffic lights, the plurality of sensors including at least one image sensor; applying a first sensor setting to the at least one image sensor to capture a first frame; applying a second sensor setting to the at least one image sensor to capture a second frame; determining the color of the traffic lights based on the sensor data of the at least one image sensor in the first frame; and controlling the ADV to drive autonomously according to the color of the traffic lights determined based on the sensor data of the at least one image sensor in the first frame and the driving environment perceived based on the sensor data of the at least one image sensor in the second frame.

[0007] In a third aspect, some embodiments of the present disclosure provide a data processing system, the data processing system including: a processor; and a memory coupled to the processor to store instructions that, when executed by the processor, cause the processor to perform operations, the operations including: perceiving a driving environment based on sensor data obtained from a plurality of sensors installed on an ADV, including detecting traffic lights, the plurality of sensors including at least one image sensor; applying a first sensor setting to the at least one image sensor to capture a first frame; applying a second sensor setting to the at least one image sensor to capture a second frame; determining the color of the traffic lights based on the sensor data of the at least one image sensor in the first frame; and controlling the ADV to drive autonomously according to the color of the traffic lights determined based on the sensor data of the at least one image sensor in the first frame and the driving environment perceived based on the sensor data of the at least one image sensor in the second frame.

[0008] In a fourth aspect, some embodiments of the present disclosure provide a computer program product, the computer program product including a computer program that, when executed by a processor, causes the processor to implement the method according to the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Embodiments of the present disclosure are illustrated by way of example and not limitation in the figures of the accompanying drawings, in which like reference numerals indicate similar elements.

[0010] Figure 1 is a block diagram showing a networked system according to one embodiment.

[0011] Figure 2 is a block diagram showing an example of an autonomous driving vehicle according to one embodiment.

[0012] Figure 3A andFigure 3B FIG. 1 is a block diagram illustrating an example of an autonomous driving system used with an autonomous vehicle according to one embodiment.

[0013] Figure 4 FIG. 2 is a block diagram illustrating an example of a control module of an autonomous driving system of an autonomous vehicle according to one embodiment.

[0014] Figure 5 FIG. 3 shows an example of pixel output relative to the light intensity of sensors of an autonomous driving system of an autonomous vehicle according to one embodiment.

[0015] Figures 6A to 6C FIG. 4 shows an example of light detection and classification of an autonomous driving system of an autonomous vehicle according to one embodiment.

[0016] Figure 7 FIG. 5 is a flowchart illustrating an example of a light detection and classification process of an autonomous driving system of an autonomous vehicle according to one embodiment. DETAILED DESCRIPTION

[0017] Various embodiments and aspects of the present disclosure will be described with reference to the details discussed below, and the accompanying drawings will illustrate the various embodiments. The following description and drawings are illustrative of the present disclosure and should not be construed as limiting the present disclosure. Many specific details are described to provide a thorough understanding of the various embodiments of the present disclosure. However, in some instances, well-known or conventional details are not described to provide a concise discussion of the embodiments of the present disclosure.

[0018] References in this specification to "one embodiment" or "an embodiment" mean that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present disclosure. The appearances of the phrase "in one embodiment" in various places in this specification do not necessarily all refer to the same embodiment.

[0019] According to some embodiments, when detecting traffic lights, the exposure time of the sensors of the ADV is reduced to, for example, the minimum exposure time or an exposure time similar to the conditions of bright light or daylight. All gains of the sensors can be set to x1. A shorter exposure time or lower gain can be applied to the sensor settings of the sensors in alternative frames. For example, one frame (Frame B) can be normal conditions. Another frame (Frame A) can be short exposure conditions (e.g., shorter exposure time or lower gain). For example, a bounding box can be defined around the traffic lights. Statistics (average value or minimum / maximum value or percentile, etc.) of red, green, and / or blue light can be extracted. The short exposure time / gain can be applied to meet the statistical thresholds of red, green, and / or blue light, which can be applied to one frame (Frame A). Thus, Frame A can be used to identify the status of traffic lights. Frame B can be used to perceive the driving environment. The ADV autonomous driving can be controlled based on both the status of traffic lights identified using Frame A and the driving environment perceived using Frame A.

[0020] According to some embodiments, the driving environment is perceived based on sensor data obtained from multiple sensors mounted on the ADV, including detecting traffic lights, wherein the multiple sensors include at least one image sensor. A first sensor setting is applied to the at least one image sensor to capture a first frame, and a second sensor setting is applied to the at least one image sensor to capture a second frame. The color of the traffic lights is determined based on the sensor data of the at least one image sensor in the first frame. The ADV is automatically driven based on the color of the traffic lights determined based on the sensor data of the at least one image sensor in the first frame and the driving environment perceived based on the sensor data of the at least one image sensor in the second frame. In this way, the ADV can identify a red traffic light signal and perceive the driving environment in a dark / cloudy environment, thereby increasing driving safety.

[0021] In one embodiment, the first sensor setting includes at least one of a first exposure time or a first gain, the second sensor setting includes at least one of a second exposure time or a second gain, and at least one of the first exposure time or the first gain is respectively less than at least one of the second exposure time or the second gain.

[0022] In one embodiment, the first sensor setting is determined based on the minimum value or a predetermined value under daylight conditions. In one embodiment, a bounding box is determined around the traffic lights in the sensor data of the at least one image sensor in the first frame, and the features of the pixels within the bounding box are extracted.

[0023] In one embodiment, the features of the pixels within the bounding box include at least one of the mean, minimum, maximum, or percentile of one of the red, green, or blue values of the pixels within the bounding box. In one embodiment, a first sensor setting is determined based on a predetermined threshold of the features of the pixels within the bounding box.

[0024] In one embodiment, when detecting traffic lights, an initial exposure time is determined in an initial setting of at least one image sensor, and it is determined whether the initial exposure time of the at least one image sensor exceeds a predetermined threshold. In response to determining that the initial setting of the at least one image sensor exceeds the predetermined threshold, the first sensor setting is applied to the at least one image sensor.

[0025] Figure 1 is a block diagram showing an autonomous driving network configuration according to an embodiment of the present disclosure. Referring Figure 1 , the network configuration 100 includes an autonomous driving vehicle (ADV) 101 communicatively coupled to one or more servers 103 to 104 via a network 102. Although one ADV is shown, multiple ADVs can be coupled to each other and / or to servers 103 to 104 via the network 102. The network 102 can be any type of network, e.g., a wired or wireless local area network (LAN), a wide area network (WAN) such as the Internet, a cellular network, a satellite network, or a combination thereof. The servers 103 to 104 can be any type of server or server cluster, such as a network or cloud server, an application server, a backend server, or a combination thereof. The servers 103 to 104 can be data analysis servers, content servers, traffic information servers, map and point of interest (MPOI) servers, or location servers, etc.

[0026] An ADV refers to a vehicle that can be configured to be in an autonomous driving mode, in which the vehicle navigates through an environment with little or no input from a driver. Such an ADV can include a sensor system having one or more sensors configured to detect information related to the vehicle's operating environment. For example, the ADV can include one or more sensors for sensing the driving environment around the ADV. The vehicle and its associated controller use the detected information to navigate through the environment. The ADV 101 can operate in a manual mode, a fully autonomous driving mode, or a partially autonomous driving mode.

[0027] In one embodiment, ADV 101 includes, but is not limited to, an autonomous driving system (ADS) 110, a vehicle control system 111, a wireless communication system 112, a user interface system 113, and a sensor system 115. ADV 101 may also include certain commonly used components included in a conventional vehicle, such as: an engine, wheels, a steering wheel, a transmission, etc., which may be controlled by the vehicle control system 111 and / or the ADS 110 using a variety of communication signals and / or commands, such as an acceleration signal or command, a deceleration signal or command, a steering signal or command, a braking signal or command, etc.

[0028] Components 110 to 115 may be communicatively coupled to each other via an interconnect, a bus, a network, or a combination thereof. For example, components 110 to 115 may be communicatively coupled to each other via a Controller Area Network (CAN) bus. The CAN bus is a vehicle bus standard designed to allow microcontrollers and devices to communicate with each other in applications without a host. It is a message-based protocol originally designed for multiplexed electrical wiring in automobiles, but is also used in many other environments.

[0029] Now referring to Figure 2 , in one embodiment, the sensor system 115 includes, but is not limited to, one or more cameras 211 (including one or more image sensors), a Global Positioning System (GPS) unit 212, an Inertial Measurement Unit (IMU) 213, a radar unit 214, and a Light Detection and Ranging (LIDAR) unit 215. The GPS unit 212 may include a transceiver that may be operable to provide information about the location of the ADV. The IMU unit 213 may sense changes in the location and orientation of the ADV based on inertial acceleration. The radar unit 214 may represent a system that uses radio signals to sense objects within the local environment of the ADV. In some embodiments, in addition to sensing objects, the radar unit 214 may additionally sense the speed and / or the forward direction of the objects. The LIDAR unit 215 may use lasers to sense objects in the environment in which the ADV is located. In addition to other system components, the LIDAR unit 215 may also include one or more laser sources, a laser scanner, and one or more detectors. The camera 211 may include one or more devices for capturing images of the environment around the ADV. For example, the camera 211 may include one or more image sensors for capturing images of the environment around the ADV. The camera 211 may be a still camera and / or a video camera. The camera may be mechanically movable, for example, by mounting the camera on a rotating and / or tilting platform.

[0030] The sensor system 115 may also include other sensors, such as: sonar sensors, infrared sensors, steering sensors, throttle sensors, brake sensors, and audio sensors (e.g., microphones). The audio sensor may be configured to collect sounds from the environment around the ADV. The steering sensor may be configured to sense the steering angle of the steering wheel, the wheels of the vehicle, or a combination thereof. The throttle sensor and the brake sensor sense the throttle position and the brake position of the vehicle, respectively. In some cases, the throttle sensor and the brake sensor may be integrated into an integrated throttle / brake sensor.

[0031] In one embodiment, the vehicle control system 111 includes, but is not limited to, a steering unit 201, a throttle unit 202 (also referred to as an acceleration unit), and a brake unit 203. The steering unit 201 is used to adjust the direction or the forward direction of the vehicle. The throttle unit 202 is used to control the speed of the electric motor or the engine, and the speed of the electric motor or the engine in turn controls the speed and acceleration of the vehicle. The brake unit 203 decelerates the vehicle by providing friction to slow down the wheels or tires of the vehicle. It should be noted that the components shown as Figure 2 may be implemented in hardware, software, or a combination thereof.

[0032] Return reference Figure 1 , the wireless communication system 112 allows communication between the ADV 101 and external systems such as devices, sensors, other vehicles, etc. For example, the wireless communication system 112 may communicate directly wirelessly with one or more devices, or communicate wirelessly via a communication network, such as communicating with servers 103 to 104 through the network 102. The wireless communication system 112 may use any cellular communication network or wireless local area network (WLAN), for example, using WiFi, to communicate with another component or system. The wireless communication system 112 may communicate directly with devices (e.g., a passenger's mobile device, a display device, a speaker within the vehicle 101) using, for example, an infrared link, Bluetooth, etc. The user interface system 113 may be part of the peripheral devices implemented within the vehicle 101, including, for example, a keyboard, a touch screen display device, a microphone, and a speaker, etc.

[0033] Some or all of the functions of the ADV 101 may be controlled or managed by the ADS 110, especially when operating in the autonomous driving mode. The ADS 110 includes the necessary hardware (e.g., processors, memories, storage devices) and software (e.g., operating systems, planning and routing programs) to receive information from the sensor system 115, the control system 111, the wireless communication system 112, and / or the user interface system 113, process the received information, plan a route or path from a starting point to a destination point, and then drive the vehicle 101 based on the planning and control information. Alternatively, the ADS 110 may be integrated with the vehicle control system 111.

[0034] For example, a user who is a passenger can specify, via a user interface for example, the starting location and destination of a trip. The ADS110 obtains trip-related data. For example, the ADV 110 can obtain location and route data from an MPOI server, which can be part of servers 103 to 104. The location server provides location services, and the MPOI server provides map services and POIs for certain locations. Alternatively, such location and MPOI information can be locally cached in the permanent storage device of the ADS 110.

[0035] When the ADV 101 moves along a route, the ADS 110 can also obtain real-time traffic information from a traffic information system or server (TIS). It should be noted that servers 103 to 104 can be operated by a third-party entity. Alternatively, the functions of servers 103 to 104 can be integrated with the ADS 110. Based on the real-time traffic information, MPOI information, location information, and real-time local environment data (such as obstacles, objects, nearby vehicles) detected or sensed by the sensor system 115, the ADS 110 can plan an optimal route and drive the vehicle 101, for example via the control system 111, to safely and efficiently reach the specified destination.

[0036] Server 103 can be a data analysis system that performs data analysis services for various customers. In one embodiment, the data analysis system 103 includes a data collector 121 and a machine learning engine 122. The data collector 121 collects driving statistics data 123 from various vehicles (ADVs or conventional vehicles driven by human drivers). The driving statistics data 123 includes information indicating the driving instructions issued (such as throttle, brake, steering instructions) and the responses of the vehicle captured by the vehicle's sensors at different time points (such as speed, acceleration, deceleration, direction). The driving statistics data 123 can also include information describing the driving environment at different time points, such as the route (including the starting location and destination location), MPOI, road conditions, weather conditions, etc.

[0037] Based on driving statistics 123, for various purposes, a machine learning engine 122 generates or trains a set of rules, algorithms, and / or prediction models 124. In one embodiment, the algorithms 124 may include algorithms or models for perceiving a driving environment (including detecting traffic lights) based on sensor data obtained from multiple sensors installed on the ADV, where the multiple sensors include at least one image sensor. The algorithms 124 may also include algorithms or models for applying a first sensor setting to at least one image sensor to capture a first frame, applying a second sensor setting to at least one image sensor to capture a second frame, determining the status of a traffic light based on sensor data of at least one image sensor in the first frame, and / or controlling the ADV for autonomous driving according to the status of the traffic light determined based on sensor data of at least one image sensor in the first frame and the driving environment perceived based on sensor data of at least one image sensor in the second frame. Then, the algorithms 124 may be uploaded to the ADV for real-time utilization during autonomous driving.

[0038] Figure 3A and Figure 3B is a block diagram showing an example of an ADV for use with an ADV according to one embodiment. The system 300 may be implemented as Figure 1 part of the ADV 101, including but not limited to the ADS 110, the control system 111, and the sensor system 115. Referring to Figures 3A to 3B , the ADS 110 includes but is not limited to a positioning module 301, a perception module 302, a prediction module 303, a decision module 304, a planning module 305, a control module 306, and a routing module 307.

[0039] Some or all of the modules 301 to 307 may be implemented in software, hardware, or a combination thereof. For example, these modules may be installed in the permanent storage device 352, loaded into the memory 351, and executed by one or more processors (not shown). It should be noted that some or all of these modules may be communicatively coupled to Figure 2 some or all of the modules of the vehicle control system 111 or integrated with them. Some of the modules 301 to 307 may be integrated together as an integrated module.

[0040] The positioning module 301 determines the current position of the ADV 300 (e.g., using the GPS unit 212) and manages any data related to the user's trip or route. The positioning module 301 (also referred to as the map and route module) manages any data related to the user's trip or route. The user can, for example, log in via the user interface and specify the starting position and destination of the trip. The positioning module 301 communicates with other components of the ADV 300, such as the map and route data 311, to obtain trip-related data. For example, the positioning module 301 can obtain location and route data from a location server and a map and POI (MPOI) server. The location server provides location services, and the MPOI server provides map services and POIs for certain locations, which can be cached as part of the map and route data 311. When the ADV 300 moves along a route, the positioning module 301 can also obtain real-time traffic information from a traffic information system or server.

[0041] Based on the sensor data provided by the sensor system 115 and the positioning information obtained by the positioning module 301, the perception module 302 determines the perception of the surrounding environment. The perception information can represent what an ordinary driver would perceive around the vehicle the driver is driving. The perception can include, for example, lane configurations in the form of objects, traffic light signals, the relative position of another vehicle, pedestrians, buildings, crosswalks, or other traffic-related signs (e.g., stop signs, yield signs), etc. The lane configuration includes information describing one or more lanes, such as, for example, the shape of the lane (e.g., straight or curved), the width of the lane, the number of lanes in the road, one-way or two-way lanes, merging or diverging lanes, exit lanes, etc.

[0042] The perception module 302 can include a computer vision system or the functionality of a computer vision system to process and analyze images captured by one or more cameras to identify objects and / or features in the ADV environment. The objects can include traffic signals, road boundaries, other vehicles, pedestrians, and / or obstacles, etc. The computer vision system can use object recognition algorithms, video tracking, and other computer vision techniques. In some embodiments, the computer vision system can map the environment, track objects, and estimate the speed of objects, etc. The perception module 302 can also detect objects based on other sensor data provided by other sensors such as radar and / or LIDAR.

[0043] For each object, the prediction module 303 predicts how the object will behave in such a situation. The prediction is performed based on the perception data that perceives the driving environment at a time point considering a set of map / route information 311 and traffic rules 312. For example, if the object is a vehicle in the opposite direction and the current driving environment includes an intersection, the prediction module 303 will predict whether the vehicle is likely to move straight ahead or turn. If the perception data indicates that there is no traffic light at the intersection, the prediction module 303 may predict that the vehicle may need to come to a complete stop before entering the intersection. If the perception data indicates that the vehicle is currently in the only lane for a left turn or the only lane for a right turn, the prediction module 303 may predict that the vehicle will be more likely to turn left or right, respectively.

[0044] For each object, the decision-making module 304 makes a decision on how to handle the object. For example, for a specific object (e.g., another vehicle in an intersecting route) and metadata describing the object (e.g., speed, direction, turning angle), the decision-making module 304 decides how to encounter the object (e.g., overtake, yield, stop, pass). The decision-making module 304 may make such a decision according to a rule set such as traffic rules or driving rules 312, which may be stored in the permanent storage device 352.

[0045] The route arrangement module 307 is configured to provide one or more routes or paths from a starting point to a destination point. For a given trip from a starting position to a destination position, e.g., a given trip received from a user, the route arrangement module 307 obtains route and map information 311 and determines all possible routes or paths from the starting position to the destination position. The route arrangement module 307 may generate a reference line in the form of a topographic map, which determines each route from the starting position to the destination position. The reference line refers to an ideal route or path that is not subject to any interference from other factors such as other vehicles, obstacles, or traffic conditions. That is, if there are no other vehicles, pedestrians, or obstacles on the road, the ADV should precisely or closely follow the reference line. Then, the topographic map is provided to the decision-making module 304 and / or the planning module 305. The decision-making module 304 and / or the planning module 305 examines all possible routes to select and modify one of the best routes according to other data provided by other modules, where the other data such as traffic conditions from the positioning module 301, the driving environment perceived by the perception module 302, and the traffic conditions predicted by the prediction module 303. According to the specific driving environment at a time point, the actual path or route for controlling the ADV may be close to or different from the reference line provided by the route arrangement module 307.

[0046] Based on the decisions for each of the perceived objects, the planning module 305 uses the reference line provided by the routing module 307 as a basis to plan a path or route and driving parameters (e.g., distance, speed, and / or turning angle) for the ADV. In other words, for a given object, the decision-making module 304 decides what to do with the object, while the planning module 305 determines how to do it. For example, for a given object, the decision-making module 304 may decide to overtake the object, and the planning module 305 may determine whether to overtake on the left or right side of the object. The planning and control data is generated by the planning module 305 and includes information describing how the vehicle 300 will move in the next movement cycle (e.g., the next route / path segment). For example, the planning and control data may instruct the vehicle 300 to move at a speed of 30 miles per hour (mph) for 10 meters and then change to the right lane at a speed of 25 mph.

[0047] Based on the planning and control data, the control module 306 controls and drives the ADV by sending appropriate commands or signals to the vehicle control system 111 according to the route or path defined by the planning and control data. The planning and control data includes sufficient information to drive the vehicle from the first point to the second point of the route or path at different time points using appropriate vehicle settings or driving parameters (e.g., throttle, brake, steering commands).

[0048] In one embodiment, the planning phase is executed in multiple planning cycles (also referred to as driving cycles), e.g., in cycles with a time interval of 100 milliseconds (ms) each. For each planning cycle or driving cycle, one or more control commands will be issued based on the planning and control data. That is, for every 100 ms, the planning module 305 plans the next route segment or path segment, e.g., including the target position and the time required for the ADV to reach the target position. Alternatively, the planning module 305 may also specify specific speeds, directions, and / or steering angles, etc. In one embodiment, the planning module 305 plans the route segment or path segment for the next predetermined time period (such as 5 seconds). For each planning cycle, the planning module 305 plans the target position for the current cycle (e.g., the next 5 seconds) based on the target position planned in the previous cycle. The control module 306 then generates one or more control commands (e.g., throttle, brake, steering control commands) based on the planning and control data for the current cycle.

[0049] Note that the decision-making module 304 and the planning module 305 can be integrated into an integrated module. The decision-making module 304 / planning module 305 can include a navigation system or the functions of a navigation system to determine the driving path of the ADV. For example, the navigation system can determine a series of speeds and forward directions for influencing the ADV to move along the following path: the path substantially avoids the perceived obstacles while making the ADV move along the lane-based path to the final destination. The destination can be set according to user input via the user interface system 113. The navigation system can dynamically update the driving path while the ADV is running. The navigation system can merge data from the GPS system and one or more maps to determine the driving path for the ADV.

[0050] Figure 4 FIG. 400 is a block diagram showing an example of the perception module and the control module of the autonomous driving system of an autonomous vehicle according to one embodiment. Refer to Figure 4 The perception module 302 includes, but is not limited to, a detection module 401 and a determination module 404. The perception module 302 is configured to perceive the driving environment based on sensor data obtained from a plurality of sensors mounted on the ADV, wherein the plurality of sensors includes at least one image sensor. The detection module 401 is configured to detect traffic lights. The control module 306 includes, but is not limited to, a first sensor setting module 402 and a second sensor setting module 403. The first sensor setting module 402 is configured to apply a first sensor setting to at least one image sensor to capture a first frame, and the second sensor setting module 403 is configured to apply a second sensor setting to at least one image sensor to capture a second frame. The determination module 404 is configured to determine the status of the traffic lights based on the sensor data of at least one image sensor in the first frame. The detection module 401 is further configured to detect the driving environment based on the sensor data of at least one image sensor in the second frame. The control module 306 is further configured to control the autonomous driving of the ADV according to the status of the traffic lights determined based on the sensor data of at least one image sensor in the first frame and the driving environment detected based on the sensor data of at least one image sensor in the second frame. For example, the control module 306 can also be configured to automatically control the brakes of the ADV according to the status of the traffic lights determined based on the sensor data of at least one image sensor in the first frame, and the control module 306 can also be configured to automatically control the wheels of the ADV according to the driving environment detected based on the sensor data of at least one image sensor in the second frame.

[0051] Figure 5FIG. 500 is an example showing pixel output 504 relative to light intensity 505 of an image sensor of an autonomous driving system of an autonomous vehicle. It is important to successfully detect traffic lights and correctly classify traffic lights (e.g., red, yellow, or green traffic light signals) for motion planning and control of the ADV. However, it is difficult to correctly identify a red traffic light signal, especially in a dark / cloudy environment. Due to color artifacts, many cases of failure to identify red traffic light signals have been reported. For example, the ADV may fail to identify a red traffic light signal due to the loss of the red traffic light signal. For another example, the ADV may incorrectly determine a red traffic light signal as a yellow traffic light signal. The failure to identify a red traffic light signal may be due to color artifacts of the image sensor of the ADV's camera.

[0052] The image sensor pixel output 504 includes a red light component (R component) 501, a green light component (G component) 502, and a blue light component (B component) 503. For example, the image sensor of the ADV's camera may use a Bayer filter, which is a color filter array (CFA) for arranging RGB color filters on a square grid of photoelectric sensors. The specific arrangement of the Bayer filter of the color filters is used in most single-chip digital image sensors in digital cameras, video cameras, and scanners to create color images.

[0053] In a dark / cloudy environment, the exposure time of the image sensor increases in a conventional sensor setting. Then, the red light intensity (R component) 501, the green light intensity (G component) 502, and the blue light intensity (B component) 503 increase according to the exposure time. However, the red light intensity (R component) has a red light intensity saturation threshold 521, which is lower than the green light intensity saturation threshold 522 or the blue light intensity saturation threshold 523.

[0054] In region 511, the red light intensity (R component) 501, the green light intensity (G component) 502, and the blue light intensity (B component) 503 increase, and the three components are in an appropriate ratio. Therefore, the pixel output 504 is red.

[0055] In region 512, when the red light intensity (R component) 501 increases to the red light intensity saturation threshold 521, the red light intensity (R component) 501 saturates and cannot become higher. The green light intensity (G component) 502 still increases. Therefore, the ratio of the R component to the G component changes, and thus, the pixel output 504 changes to orange in region 512.

[0056] In region 513, green light intensity (G component) 502 is saturated when it reaches green light intensity saturation threshold 522. Therefore, the ratio of R component 501 to G component 502 is close to 1, and B component 503 is low. Therefore, pixel output 504 appears yellow.

[0057] In region 514, when all three components are saturated, pixel output 504 appears white.

[0058] like Figure 5 As shown in , in dark / cloudy environments, the increased exposure time of the image sensor in a conventional sensor setup may cause artifacts, which may lead to a failure in recognizing a red traffic light signal.

[0059] Figures 6A to 6C An example 600 of light detection and classification for an autonomous driving system of an autonomous vehicle is shown, according to one embodiment. Figure 6A An example of the process of light detection and classification is shown. Figure 6B Shows Figure 6A 602 in frame A. FIG. 604 is a magnified view of portion 610a of frame A 602. Figure 6C Shows Figure 6A FIG. 6 is an enlarged view of portion 610b of frame B 603. Figures 6A to 6C , the ADV may perceive the driving environment based on sensor data obtained from a plurality of sensors mounted on the ADV, wherein the plurality of sensors include one or more image sensors. Each of the image sensors may have sensor settings. For example, the sensor settings may include exposure time, gain, etc.

[0060] At the initial time T1, the ADV may detect one or more traffic lights 620 based on sensor data obtained from multiple sensors including one or more image sensors. One or more cameras may be installed on the ADV, and each of the one or more cameras may include an image sensor. In a dark / cloudy environment, the initial exposure time in the initial sensor setting of the one or more image sensors may be increased (e.g., the exposure time at time T1) to capture the driving environment.

[0061] In one embodiment, an initial exposure time in an initial sensor setting of one or more image sensors may be determined when detecting traffic light 620. It may be determined whether the initial exposure time of the one or more image sensors exceeds a predetermined threshold. For example, the predetermined threshold may be based on a red light intensity saturation threshold (e.g., Figure 5 521) in the determination. Figure 5As discussed, when the initial exposure time in the initial settings of one or more image sensors exceeds a predetermined threshold, the ADV may not be able to recognize red traffic lights. Therefore, in order to correctly recognize the red traffic light signal, the exposure time can be reduced to the minimum exposure time or an exposure time similar to the conditions in bright sunlight or daylight.

[0062] At time T2, the first sensor settings can be applied to one or more image sensors in the first frame (i.e., "Frame A" 610a). The first sensor settings can include a first exposure time and / or a first gain. For example, the first exposure time can be a reduced exposure time, such as the minimum exposure time or an exposure time similar to the conditions in bright sunlight or daylight. As another example, all gains can be set to x1 or the minimum gain.

[0063] As Figure 6B shown, in the enlarged view of part 610a of "Frame A" 602, for example, due to the reduced exposure time and / or reduced gain, the red traffic light 620 can appear red. The determination module 404 in the perception module 302 can determine the color of the traffic light 620 as red, and thus the red traffic light 620 can be recognized. However, it is difficult to recognize other driving environments in the first frame "Frame A" 602, such as lane configurations, obstacles, etc.

[0064] At time T3, the second sensor settings can be applied to one or more image sensors in the second frame (i.e., "Frame B" 610b). The second sensor settings can include a second exposure time and / or a second gain. The second sensor settings can be the normal sensor settings in a dark / overcast environment. For example, the second exposure time can be an exposure time longer than the exposure time in bright sunlight or daylight. As another example, all gains can be set to greater than 1x or the minimum gain.

[0065] As Figure 6C shown, in the enlarged view of part 610b of "Frame B" 603, for example, due to the increase in the exposure time combined Figure 5 with that discussed, the red traffic light 620 can appear orange or yellow. However, the perception module 302 can perceive the driving environment based on the sensor data of one or more sensors in the second frame "Frame B" 603. The ADV can recognize other driving environments in the second frame "Frame B" 603, such as lane configurations, obstacles.

[0066] The control module 306 of the ADV can apply the first sensor settings and the second sensor settings to alternate frames. In one embodiment, the ADV can apply the first sensor settings to one or more sensors to capture the first frame, and apply the second sensor settings to one or more sensors to capture the second frame, and repeat.

[0067] In one embodiment, a bounding box 622 around a traffic light 620 in sensor data of one or more sensors in a first frame "Frame A" 602 may be determined. Each traffic light 620 may have a corresponding bounding box 622. Features of pixels within the bounding box 622 may be extracted. For example, features of pixels within the bounding box 622 may include an average value, a minimum value, a maximum value, or a percentile of one of the red (R component), green (G component), or blue (B component) of the pixels within the bounding box 622. As an example, the feature may be the average value of the red (R component).

[0068] In one embodiment, a first sensor setting may be determined based on a predetermined threshold of features of pixels within the bounding box. For example, a first exposure time may be set to meet a predetermined threshold of an average value, a minimum value, a maximum value, or a percentile of one of the red (R component), green (G component), or blue (B component) of the pixels within the bounding box 622. As an example, the first exposure time may be set to meet a predetermined threshold of the average value of the red (R component) of the pixels within the bounding box 622 in the first frame "Frame A" 602. As another example, the first exposure time may be set to meet a predetermined threshold of the maximum value of the red (R component) of the pixels within the bounding box 622 in the first frame "Frame A" 602.

[0069] The ADV may alternately apply a first sensor setting and a second sensor setting. The ADV may apply the first sensor setting to capture a plurality of first frames, or apply the second sensor setting to capture a plurality of second frames, depending on light conditions and / or the environment. In one embodiment, the ADV may apply the first sensor setting to one or more image sensors to capture a plurality of first frames, and apply the second sensor setting to one or more image sensors to capture a second frame. In one embodiment, the ADV may apply the first sensor setting to one or more image sensors to capture a first frame, and apply the second sensor setting to one or more image sensors to capture a plurality of second frames. In one embodiment, the ADV may apply the first sensor setting to one or more image sensors to capture a plurality of first frames, and apply the second sensor setting to one or more image sensors to capture a plurality of second frames.

[0070] The ADV autonomous driving can be controlled based on the color of the traffic light determined according to the sensor data of one or more image sensors in the first frame and the driving environment perceived based on the sensor data of one or more image sensors in the second frame. For example, the brakes of the ADV can be applied to stop the ADV in response to determining that the color in the traffic light is red based on the sensor data of one or more image sensors in the first frame. For example, the wheels of the ADV can be rotated to change the trajectory of the ADV in response to an obstacle perceived based on the sensor data of one or more image sensors in the second frame. In this way, the ADV can identify the red traffic light signal and perceive the driving environment in a dark / cloudy environment, thereby increasing driving safety.

[0071] Figure 7 FIG. is a flowchart illustrating an example of a process of optical detection and classification of an autonomous driving system of an autonomous vehicle according to one embodiment. Process 700 can be executed by processing logic that can include software, hardware, or a combination thereof. For example, process 700 can be executed by the perception module 302 and / or the control module 306. Refer to Figure 7 , in operation 701, the processing logic perceives the driving environment around the ADV based on the sensor data obtained from a plurality of sensors installed on the ADV, including detecting traffic lights, where the plurality of sensors includes at least one image sensor. In operation 702, the processing logic applies a first sensor setting to the at least one image sensor to capture a first frame. In operation 703, the processing logic applies a second sensor setting to the at least one image sensor to capture a second frame. In operation 704, the processing logic determines the color of the traffic light based on the sensor data of the at least one image sensor in the first frame. In operation 705, the processing logic controls the ADV autonomous driving according to the color of the traffic light determined based on the sensor data of the at least one image sensor in the first frame and the driving environment perceived based on the sensor data of the at least one image sensor in the second frame. Through this process, the ADV can identify the red traffic light signal and perceive the driving environment in a dark / cloudy environment, thereby increasing driving safety.

[0072] Note that some or all of the components shown and described above may be implemented in software, hardware, or a combination thereof. For example, such components may be implemented as software installed and stored in a permanent storage device, which may be loaded into and executed in a memory by a processor (not shown) to implement the processes or operations described throughout this application. Alternatively, such components may be implemented as executable code programmed or embedded in dedicated hardware, such as an integrated circuit (e.g., an application specific integrated circuit or ASIC), a digital signal processor (DSP), or a field programmable gate array (FPGA), which may be accessed via corresponding drivers from an application and / or an operating system. Additionally, such components may be implemented as specific hardware logic in a processor or processor core, as part of an instruction set that may be accessed by software components via one or more specific instructions.

[0073] Some portions of the foregoing detailed description have been presented in terms of algorithms and symbolic representations of operations on data bits within a computer memory. These algorithmic descriptions and representations are the means used by those skilled in the data processing arts to most effectively convey the substance of their work to others skilled in the art. In this context, an algorithm is usually conceived of as a self-consistent sequence of operations leading to a desired result. These operations are those requiring physical manipulation of physical quantities.

[0074] However, it should be borne in mind that all such and similar terms are to be associated with appropriate physical quantities and are merely convenient labels applied to these quantities. Unless explicitly stated otherwise in the foregoing discussion, it should be understood that throughout the specification, discussions using terms, such as those set forth in the appended claims, refer to the actions and processes of a computer system or similar electronic computing device that manipulates data represented as physical (electronic) quantities within the registers and memories of the computer system and transforms that data into other data similarly represented as physical quantities within the memories or registers of the computer system or other such information storage, transmission, or display devices.

[0075] Embodiments of the present disclosure also relate to an apparatus for performing the operations herein. Such a computer program is stored in a non-transitory computer-readable medium. A machine-readable medium includes any mechanism for storing information in a form readable by a machine, e.g., a computer. For example, a machine-readable (e.g., computer-readable) medium includes a machine (e.g., computer) readable storage medium (e.g., read only memory (“ROM”), random access memory (“RAM”), magnetic disk storage media, optical storage media, flash memory devices).

[0076] The processes or methods depicted in the foregoing figures may be performed by processing logic, which includes hardware (e.g., circuitry, dedicated logic, etc.), software (e.g., embodied on a non-transitory computer-readable medium), or a combination of both. Although the processes or methods have been described above in terms of some sequential operations, it should be understood that some of the operations may be performed in a different order. Additionally, some operations may be performed in parallel rather than sequentially.

[0077] Embodiments of the present disclosure have not been described with reference to any particular programming language. It should be recognized that a variety of programming languages may be used to implement the teachings of the embodiments of the present disclosure as described herein.

[0078] In the foregoing specification, embodiments of the present disclosure have been described with reference to specific exemplary embodiments of the present disclosure. It will be apparent that various modifications may be made to the invention without departing from the broader spirit and scope of the present disclosure as set forth in the appended claims. Accordingly, the specification and drawings are to be interpreted in an illustrative rather than a restrictive sense.

Claims

1. A computer-implemented method for operating an autonomous driving vehicle (ADV), the method comprising: Perceiving a driving environment based on sensor data obtained from a plurality of sensors mounted on the ADV, including detecting a traffic light at an initial time, the plurality of sensors including at least one image sensor; Applying a first sensor setting to the at least one image sensor to capture a first frame; Applying a second sensor setting to the at least one image sensor to capture a second frame; Determining a color of the traffic light based on sensor data of the at least one image sensor in the first frame; And Controlling the ADV to drive autonomously according to the color of the traffic light determined based on sensor data of the at least one image sensor in the first frame and the driving environment perceived based on sensor data of the at least one image sensor in the second frame; Wherein, when the traffic light is detected, determining an initial exposure time at the initial time in an initial sensor setting of the at least one image sensor; determining whether the initial exposure time of the at least one image sensor exceeds a predetermined threshold; in response to determining that the initial exposure time of the at least one image sensor exceeds the predetermined threshold, applying the first sensor setting to the at least one image sensor; the predetermined threshold is determined based on a red light intensity saturation threshold; The first sensor setting includes at least one of a first exposure time or a first gain, wherein the second sensor setting includes at least one of a second exposure time or a second gain, and wherein at least one of the first exposure time or the first gain is respectively less than at least one of the second exposure time or the second gain.

2. The method according to claim 1, further comprising: Determining a bounding box around the traffic light in the sensor data of the at least one image sensor in the first frame; And Extracting features of pixels within the bounding box.

3. The method according to claim 2, wherein, The features of the pixels within the bounding box include at least one of an average value, a minimum value, a maximum value, or a percentile of one of red, green, or blue of the pixels within the bounding box.

4. The method according to claim 2, wherein Determining the first sensor setting based on a predetermined threshold of the features of the pixels within the bounding box.

5. A non-transitory machine-readable medium storing instructions that, when executed by a processor, cause the processor to perform operations, the operations including: Perceiving a driving environment based on sensor data obtained from a plurality of sensors mounted on an ADV, including detecting a traffic light at an initial time, the plurality of sensors including at least one image sensor; Applying a first sensor setting to the at least one image sensor to capture a first frame; Applying a second sensor setting to the at least one image sensor to capture a second frame; Determining a color of the traffic light based on sensor data of the at least one image sensor in the first frame; And Control the ADV autonomous driving according to the color of the traffic light determined based on the sensor data of the at least one image sensor in the first frame and the driving environment perceived based on the sensor data of the at least one image sensor in the second frame; Wherein, when the traffic light is detected, determine the initial exposure time at the initial time in the initial sensor settings of the at least one image sensor; determine whether the initial exposure time of the at least one image sensor exceeds a predetermined threshold; in response to determining that the initial exposure time of the at least one image sensor exceeds the predetermined threshold, apply the first sensor setting to the at least one image sensor; the predetermined threshold is determined based on the red light intensity saturation threshold; The first sensor setting includes at least one of a first exposure time or a first gain, wherein the second sensor setting includes at least one of a second exposure time or a second gain, and wherein at least one of the first exposure time or the first gain is respectively less than at least one of the second exposure time or the second gain.

6. The non-transitory machine-readable medium according to claim 5, wherein, The operation further includes: Determine a bounding box around the traffic light in the sensor data of the at least one image sensor in the first frame; and Extract the features of the pixels within the bounding box.

7. The non-transitory machine-readable medium according to claim 6, wherein, The features of the pixels within the bounding box include at least one of the average value, minimum value, maximum value or percentile of one of red, green or blue of the pixels within the bounding box.

8. The non-transitory machine-readable medium according to claim 6, wherein Determine the first sensor setting based on a predetermined threshold of the features of the pixels within the bounding box.

9. A data processing system, comprising: A processor; And A memory coupled to the processor to store instructions that, when executed by the processor, cause the processor to perform operations, the operations including: Perceive the driving environment based on sensor data obtained from multiple sensors installed on the ADV, including detecting a traffic light at an initial time, the multiple sensors including at least one image sensor; Apply a first sensor setting to the at least one image sensor to capture a first frame; Apply a second sensor setting to the at least one image sensor to capture a second frame; Determine the color of the traffic light based on the sensor data of the at least one image sensor in the first frame; and Control the ADV autonomous driving according to the color of the traffic light determined based on the sensor data of the at least one image sensor in the first frame and the driving environment perceived based on the sensor data of the at least one image sensor in the second frame; Among them, when the traffic light is detected, determining an initial exposure time of the at least one image sensor at the initial time in the initial sensor settings; determining whether the initial exposure time of the at least one image sensor exceeds a predetermined threshold; in response to determining that the initial exposure time of the at least one image sensor exceeds the predetermined threshold, applying the first sensor settings to the at least one image sensor; the predetermined threshold is determined based on a red light intensity saturation threshold; The first sensor settings include at least one of a first exposure time or a first gain, wherein the second sensor settings include at least one of a second exposure time or a second gain, and wherein at least one of the first exposure time or the first gain is respectively less than at least one of the second exposure time or the second gain.

10. The data processing system according to claim 9, wherein, The operation further includes: Determining a bounding box around the traffic light in the sensor data of the at least one image sensor in the first frame; and Extracting features of pixels within the bounding box.

11. The data processing system according to claim 10, wherein, The features of the pixels within the bounding box include at least one of an average value, a minimum value, a maximum value, or a percentile of one of red, green, or blue of the pixels within the bounding box.

12. The data processing system according to claim 10, wherein, Determining the first sensor settings based on a predetermined threshold of the features of the pixels within the bounding box.

13. A computer program product including a computer program, which when executed by a processor causes the processor to implement the method according to any one of claims 1 to 4.

Images