Method for detecting and classifying objects in road traffic

Patent metadata

Field	Value
Publication number	US-12482240-B2
Application number	US-202118042750-A
Country	US
Kind code	B2
Filing date	Jul 26, 2021
Priority date	Sep 16, 2020
Publication date	Nov 25, 2025
Grant date	Nov 25, 2025

Abstract

A method (40) for detecting and classifying at least one object in road traffic is disclosed. Sensor data from a sensor are first of all provided. At least one object is detected and classified on the basis of the sensor data using a neural network. The object is additionally detected and classified on the basis of the sensor data using a symbolic monitoring algorithm. A check is carried out in order to determine whether the neural network and the symbolic monitoring algorithm provide consistent results with respect to the detection and classification of the object.

First claim

The invention claimed is:

1 . A method for detecting and classifying at least one object in road traffic, comprising:
providing sensor data from a sensor; detecting and classifying at least one object based on the sensor data using a neural network; detecting and classifying the at least one object based on the sensor data using a symbolic monitoring algorithm; checking whether the neural network and the symbolic monitoring algorithm provide consistent results with respect to the detection and classification of the at least one object; and before the consistency check, checking with an additional symbolic monitoring algorithm whether the at least one object detected and classified by the neural network has predefined properties.

2 . The method according to claim 1 , wherein the symbolic monitoring algorithm and/or the additional symbolic monitoring algorithm comprises a plurality of symbolic sub-algorithms.

3 . The method according to claim 1 , wherein the sensor data are provided at a predefined frequency and the consistency check is carried out at predefined time intervals.

4 . The method according to claim 1 , wherein results of the detection and classification of the neural network and the symbolic monitoring algorithm are stored in a memory, the method further comprising:
creating a set of statistics on the results of a plurality of consistency checks within a predefined time period.

5 . The method according to claim 4 , wherein the set of statistics contains information on how often the at least one object was detected by the neural network and by the symbolic monitoring algorithm and how often results of the neural network were validated by the symbolic monitoring algorithm with regard to detection and classification.

6 . The method according to claim 1 , wherein the sensor is a camera which provides images as sensor data.

7 . The method according to claim 1 , wherein the sensor is a component of an autonomous motor vehicle.

8 . The method according to claim 7 , wherein an object that is considered to be detected and classified is taken into account within a framework of an automatic control system of the autonomous motor vehicle.

9 . A system which is configured to carry out the method according to claim 1 .

10 . The method according to claim 1 , wherein a computer program comprises instructions which, when the computer program is executed by a computer, cause the computer to carry out the method.

11 . A non-transitory machine-readable storage medium on which the computer program according to claim 10 is stored.

12 . A method for detecting and classifying at least one object in road traffic, comprising:
providing sensor data from a sensor; detecting and classifying at least one object based on the sensor data using a neural network; detecting and classifying the at least one object based on the sensor data using a symbolic monitoring algorithm; and checking whether the neural network and the symbolic monitoring algorithm provide consistent results with respect to the detection and classification of the at least one object, wherein the at least one object is considered to be detected and classified when there is a predefined number of consistent results, and wherein the predefined number of consistent results is determined depending on a distance of the at least one object from the sensor.

Assignees

• Bosch Gmbh Robert

Inventors

• Lutz Robert
• Lindner Frank
• Hundsdoerfer Daniel
• Kausch Jonathan
• Novak Libor

Classifications

• G06V20/58

  Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads · CPC title

• G06V10/776 Primary

  Validation; Performance evaluation · CPC title

• G06V10/82

  using neural networks · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title

• G06F18/254

  of classification results, e.g. of results related to same input data · CPC title