US-12620127-B2 - Providing a prediction of a roll angle of a motorcycle

US12620127B2US 12620127 B2US12620127 B2US 12620127B2US-12620127-B2

Abstract

The present invention detects a stage of a motorcycle by use of a camera which takes one or more images of an environment of the vehicle. There is a known angular relationship between the field of view of the camera and predefined par or a vertical axis of the motorcycle so that signatures of one or more images are generated. The signatures are used to search multiple groups of concept data structures, for one or more matching concept data structures, wherein the different groups of concept data structures represent images virtually acquired at different roll angles. A roll angle of the motorcycle is determined based on an outcome of the searching and the known angular relationship.

Inventors

Adam HAREL
Karina ODINAEV

Assignees

AUTOBRAINS TECHNOLOGIES LTD

Dates

Publication Date: 20260505
Application Date: 20230620

Claims (15)

1 . A method for detecting a state of a motorcycle, the method comprises: obtaining, by a camera of the motorcycle, one or images of an environment of the vehicle; wherein there is a known angular relationship between the field of view of the camera and predefined part or a vertical axis of the motorcycle; generating signatures of the one or more images; searching, using the signatures, multiple groups of concept data structures, for one or more matching concept data structures; wherein different groups of concept data structures represent images virtually acquired at different roll angles; wherein the different groups of concept data structures are generated by a training process that guarantees that the groups of concept data structures are roll angle sensitive; and determining a roll angle of the motorcycle based on an outcome of the searching and the known angular relationship.
2 . The method according to claim 1 wherein when the outcome provides a single matching concept data structure then the determining of the roll angle comprises adopting a roll angle associated with the single matching concept data structure.
3 . The method according to claim 1 wherein when the outcome provides a plurality of matching concept data structures then the determining of the roll angle comprises calculating the roll angle based on two or more roll angles associated with two or more of the plurality of matching concept data structures.
4 . The method according to claim 1 wherein when the outcome provides a plurality of matching concept data structures then the determining of the roll angle comprises selecting a roll angle associated with one of the plurality of matching concept data structures.
5 . The method according to claim 4 wherein the one of the plurality of matching concept data structures is a best matching concept data structure of the plurality of matching concept data structures.
6 . The method according to claim 1 , wherein the training process is implemented by convolution neural networks.
7 . The method according to claim 1 wherein the different groups of concept data structures are generated by a process that comprising removing concept data structures that represents objects that exhibit roll angle ambiguity.
8 . A method for detecting a state of a motorcycle, the method comprises: obtaining, by a camera of the motorcycle, one or images of an environment of the vehicle; wherein there is a known angular relationship between the field of view of the camera and predefined part or a vertical axis of the motorcycle; generating signatures of the one or more images; searching, using the signatures, multiple groups of concept data structures, for one or more matching concept data structures; wherein different groups of concept data structures represent images virtually acquired at different roll angles; and determining a roll angle of the motorcycle based on an outcome of the searching and the known angular relationship; wherein the multiple groups of concept data structures are generated by applying an unsupervised machine learning process on (a) images acquired by cameras of four-wheeled vehicles, and (b) roll axis rotated versions of the images acquired by cameras of four-wheeled vehicles.
9 . A non-transitory computer readable medium for detecting a state of a motorcycle, the non-transitory computer readable medium stores instructions for: obtaining, by a camera of the motorcycle, one or images of an environment of the vehicle; wherein there is a known angular relationship between the field of view of the camera and predefined part or a vertical axis of the motorcycle; generating signatures of the one or more images; searching, using the signatures, multiple groups of concept data structures, for one or more matching concept data structures; wherein different groups of concept data structures represent images virtually acquired at different roll angles; and determining a roll angle of the motorcycle based on an outcome of the searching and the known angular relationship; and wherein when the outcome provides a plurality of matching concept data structures then the determining of the roll angle comprises selecting a roll angle associated with one of the plurality of matching concept data structures.
10 . The non-transitory computer readable medium according to claim 9 wherein when the outcome provides a single matching concept data structure then the determining of the roll angle comprises adopting a roll angle associated with the single matching concept data structure.
11 . The non-transitory computer readable medium according to claim 9 wherein the one of the plurality of matching concept data structures is a best matching concept data structure of the plurality of matching concept data structures.
12 . The non-transitory computer readable medium according to claim 9 wherein the multiple groups of concept data structures are generated by applying an unsupervised machine learning process on (a) images acquired by cameras of four-wheeled vehicles, and (b) roll axis rotated versions of the images acquired by cameras of four-wheeled vehicles.
13 . The non-transitory computer readable medium according to claim 9 wherein the different groups of concept data structures are generated by a training process that guarantees that the groups of concept data structures are roll angle sensitive.
14 . The non-transitory computer readable medium according to claim 9 wherein the training process is implemented by convolution neural networks.
15 . The non-transitory computer readable medium according to claim 9 wherein the different groups of concept data structures are generated by a process that comprising removing concept data structures that represents objects that exhibit roll angle ambiguity.

Description

BACKGROUND The roll angle of a motorcycle is one of the most crucial safety related aspects of its performance. An incorrect roll angle may result in an accident. There is a growing need to monitor the roll angle of a motorcycle. SUMMARY A method, system and non-transitory computer readable medium for providing a prediction of a roll angle of a motorcycle. BRIEF DESCRIPTION OF THE DRAWINGS The embodiments of the disclosure will be understood and appreciated more fully from the following detailed description, taken in conjunction with the drawings in which: FIG. 1A illustrates an example of a method; FIG. 1B illustrates an example of a signature; FIG. 1C illustrates an example of a dimension expansion process; FIG. 1D illustrates an example of a merge operation; FIG. 1E illustrates an example of hybrid process; FIG. 1F illustrates an example of a method; FIG. 1G illustrates an example of a method; FIG. 1H illustrates an example of a method; FIG. 1I illustrates an example of a method; FIG. 1J illustrates an example of a method; FIG. 1K illustrates an example of a method; FIG. 1L illustrates an example of a method; FIG. 1M illustrates an example of a system; FIG. 1N is a partly-pictorial, partly-block diagram illustration of an exemplary obstacle detection and mapping system, constructed and operative in accordance with embodiments described herein; FIG. 1O illustrates an example of a method; FIG. 1P illustrates an example of a method; FIG. 1Q illustrates an example of a method; FIG. 2 illustrates an example of a method; FIG. 3 illustrates examples of two different images acquired at two different roll angles; and FIG. 4 illustrates an example of a motorcycle. DESCRIPTION OF EXAMPLE EMBODIMENTS The specification and/or drawings may refer to an image. An image is an example of a media unit. Any reference to an image may be applied mutatis mutandis to a media unit. A media unit may be an example of sensed information. Any reference to a media unit may be applied mutatis mutandis to a natural signal such as but not limited to signal generated by nature, signal representing human behavior, signal representing operations related to the stock market, a medical signal, and the like. Any reference to a media unit may be applied mutatis mutandis to sensed information. The sensed information may be sensed by any type of sensors—such as a visual light camera, or a sensor that may sense infrared, radar imagery, ultrasound, electro-optics, radiography, LIDAR (light detection and ranging), etc. The specification and/or drawings may refer to a processor. The processor may be a processing circuitry. The processing circuitry may be implemented as a central processing unit (CPU), and/or one or more other integrated circuits such as application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), full-custom integrated circuits, etc., or a combination of such integrated circuits. Any combination of any steps of any method illustrated in the specification and/or drawings may be provided. Any combination of any subject matter of any of claims may be provided. Any combinations of systems, units, components, processors, sensors, illustrated in the specification and/or drawings may be provided. There may be provided a system, a method and a non-transitory computer readable medium for estimating a state of a motorcycle. The method may determine the roll angle of the motorcycle. The determining may be executed in real time—for example per each frame, multiple times a second, and the like. The roll angle may be used for one or more purposes such as performing an ADAS operation such as alerting the driver, estimate the risk (for example for insurance purposes) associated with the driving of the motorcycle, providing feedback to a perception engine of the motorcycle, finding misalignments between the camera and the predefined part of the motorcycle, and the like. Regarding the perception engine—the perception engine may operate while using assumptions regarding the environment of the motorcycle, the orientation of the camera (for example roll angle). The suggested method may provide information about the roll angle and may improve the accuracy of the perception module. FIG. 2 is an example of a method 100 for determining a state of a motorcycle. Method 100 may include steps 110, 120, 130 and 140. Step 110 may include obtaining, by a camera of the motorcycle, one or images of an environment of the vehicle. The one or more images are taken when there is a known angular relationship between the field of view of the camera and predefined part or a vertical axis or a vertical axis of the motorcycle. The predefined part may be any part of the frame, the steering damper, gas tank, engine, front fork, seat, and the like. The angular relationship be include the roll angle but may also include at least one of a yaw angle or a pitch angle. The known angular relationship may relate to the optical axis of the camera, to a long