EP4587856A1 - Verfahren zum trainieren eines neuronalen netzes zur detektion eines objekts und verfahren zur detektion eines objekts mittels eines neuronalen netzes - Google Patents
Verfahren zum trainieren eines neuronalen netzes zur detektion eines objekts und verfahren zur detektion eines objekts mittels eines neuronalen netzesInfo
- Publication number
- EP4587856A1 EP4587856A1 EP23751280.1A EP23751280A EP4587856A1 EP 4587856 A1 EP4587856 A1 EP 4587856A1 EP 23751280 A EP23751280 A EP 23751280A EP 4587856 A1 EP4587856 A1 EP 4587856A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- neural network
- radar
- test object
- mixed
- spectrum
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/417—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section involving the use of neural networks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/86—Combinations of radar systems with non-radar systems, e.g. sonar, direction finder
- G01S13/867—Combination of radar systems with cameras
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10048—Infrared image
Definitions
- the invention relates to a method for training a neural network to detect an object. Training data is generated and fed to the neural network. The invention also relates to a method for detecting an object using a neural network.
- a camera and a radar are used together.
- the invention is based on the object of developing a method for training a neural network to detect an object, as well as a method for detecting an object using a neural network.
- the mixed spectrum contains information about a distance, an azimuth angle, an elevation angle and a radial velocity of the test object.
- the radial velocity is determined via a frequency shift between the transmitted radar signal and the reflected radar signal.
- the frequency shift in question results from the Doppler effect in moving objects.
- markings are attached to the test object in such a way that the markings are visible in the images created.
- a six-dimensional pose of the test object is computable, which includes a position of the test object and an orientation of the test object.
- a method for detecting an object using a neural network is also proposed, with training data previously being supplied to the neural network.
- the training data was fed to the neural network using the method according to the invention for training a neural network.
- a radar signal is sent out by a radar sensor and a radar signal reflected by the object is received.
- the transmitted radar signal and the received radar signal are mixed into a mixed signal, and a mixed spectrum of the mixed signal is calculated.
- Input data containing the mixed spectrum is fed to the neural network.
- the input data is processed in the neural network.
- the object and a position of the object are detected by the neural network.
- the neural network outputs an object class of the detected object and the detected position of the object as output data.
- a unique signature in the mixed spectrum enables robust classification.
- the neural network is designed as a convolutional network which has an input layer, an output layer and a plurality of convolution layers. This is done from one layer to the next layer performed a folding operation.
- the layers are arranged in series one behind the other and linked to one another using mathematical folding operations.
- a convolution operation is carried out from one layer to the next layer.
- the calculated mixed spectrum includes at least a distance and an azimuth angle of a radar measurement.
- the neural network is fed initial input data, which contains the distance of the radar measurement.
- the neural network is fed second input data, which contains the azimuth angle of the radar measurement.
- the first input data and the second input data represent a first complex image made of complex data.
- the first complex image thus comprises two simple images which contain magnitude and phase.
- Figure 1 a schematic representation of an arrangement for obtaining training data
- Figure 4 a schematic representation of output data from a neural network.
- Figure 1 shows a schematic representation of an arrangement for obtaining training data for a neural network 7.
- the arrangement has a measuring area 40 and a radar area 42.
- the measuring area 40 and the radar area 42 largely overlap.
- a test object, not shown here, is located within the measuring range 40 and within the radar range 42.
- the radar device 25 has a 2-D MIMO (Multiple Input Multiple Output) antenna array.
- the emitted radar signal has FMCW modulation (Frequency-Modulated Continuous Wave).
- the calculated mixed spectrum is therefore four-dimensional and contains information about a distance, an azimuth angle, an elevation angle and a radial speed of the test object from which the radar signal is reflected.
- the first partial spectrum contains a first radar image with information about an amount of the distance and the azimuth angle of the test object.
- the first subspectrum also contains a second radar image with information about a phase of the range and azimuth angle of the test object.
- the second sub-spectrum contains a third radar image with information about an amount of distance and the radial velocity of the test object.
- the second subspectrum also contains a fourth radar image with information about a phase of the distance and the radial velocity of the test object.
- the radar images are available in polar coordinates.
- the arrangement further comprises a digital computer 32 and a processing unit 34.
- the cameras 21 are connected to the processing unit 34 and transmit recorded images to the processing unit 34.
- the radar device 25 is also connected to the processing unit 34 and transmits data to the processing unit 34.
- the processing unit 34 is connected to the digital computer 32 and transmits data to the digital computer 32.
- geometric dimensions of the test object are recorded.
- the length, width and height of the test object are measured.
- Markings are also attached to the test object.
- the markings in question are designed as infrared markers.
- the cameras 21 are designed as infrared cameras. The markings are attached to the test object in such a way that the markings are visible in recordings later generated by the cameras 21.
- the acquisition of the training data for the neural network 7 using the selected test object takes place during a previously defined period of time.
- the test object is moved in an area which lies within the measuring range 40 and within the radar range 42. If necessary, the test object moves independently in the said area during the period.
- the cameras produce 21 images of the test object.
- a pose of the test object is calculated from the recordings.
- the pose in question is six-dimensional and includes a position of the test object and an orientation of the test object.
- Occupancy maps are generated from the previously recorded geometric dimensions of the test object and the images created.
- the calculated poses are integrated into the occupancy maps.
- the occupancy cards are assigned to the object class of the selected test object.
- the occupancy maps are first generated in Cartesian coordinates, and the Cartesian coordinates are then transformed into polar coordinates.
- a radar signal is simultaneously transmitted by the radar device 25 and a radar signal reflected by the test object is received.
- the transmitted radar signal and the received radar signal are mixed to form a mixed signal.
- a complex four-dimensional mixed spectrum of the mixed signal is also calculated.
- the complex four-dimensional mixed spectrum becomes a first complex two-dimensional partial spectrum and a second complex two-dimensional one Partial spectrum calculated.
- the partial spectra contain radar images which are available in polar coordinates.
- the occupancy maps and the partial spectra are then merged to form training data.
- the training data is assigned to the respective object class of the selected test object.
- the training data obtained in this way is fed to the neural network 7.
- the method steps described for obtaining the training data for the neural network 7 are repeated for further test objects from further object classes. Test objects are selected from other object classes. Furthermore, the method steps described for obtaining the training data for the neural network 7 are carried out once without a real test object, but with a free space. The occupancy maps and the training data are assigned to the respective object class or the free space.
- FIG. 2 shows a schematic representation of a neural network 7.
- the neural network 7 is designed as a convolutional network.
- the neural network 7 in the present case has an input layer 6, a first convolution layer 11, a second convolution layer 12, a third convolution layer 13, a fourth convolution layer 14, a fifth convolution layer 15, a sixth convolution layer 16, a seventh convolution layer 17 and an output layer 9.
- Input data 1, 2, 3, 4 are fed to the input layer e of the neural network 7.
- the input layer e, the convolution layers 11, 12, 13, 14, 15, 16, 17 and the output layer 9 are arranged in series one after the other.
- a convolution operation is carried out from one layer to the next layer.
- Output data 51, 52, 53, 54 are output from the output layer 9 of the neural network 7.
- the input data 1, 2, 3, 4 are processed in the neural network 7.
- a folding operation is carried out from one layer to the next layer.
- the neural network 7 detects the object and a position of the object.
- an object class of the object is also detected by the neural network 7.
- Output data 51, 52, 53, 54 are output from the output layer 9 of the neural network 7.
- Figure 4 shows a schematic representation of output data 51, 52, 53, 54 of the neural network 7.
- the first output data 51 is assigned to an object from a first object class, for example a person.
- the first output data 51 contains the detected position of the object.
- the first output data 51 represents a two-dimensional matrix of individual pixels. In the present case, the first output data 51 has a size of 64x64 pixels.
- the second output data 52 is assigned to an object from a second object class, for example a forklift.
- the second output data 52 contains the detected position of the object.
- the second output data 52 represents a two-dimensional matrix of individual pixels. In the present case, the second output data 52 has a size of 64x64 pixels.
- the third output data 53 is assigned to an object from a third object class, for example an autonomous transport vehicle.
- the third output data 53 contains the detected position of the object.
- the third output data 53 represents a two-dimensional matrix of individual pixels. In the present case, the third output data 53 has a size of 64x64 pixels.
Landscapes
- Engineering & Computer Science (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Computer Networks & Wireless Communication (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Health & Medical Sciences (AREA)
- Image Analysis (AREA)
Abstract
Description
Claims
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| DE102022003398 | 2022-09-15 | ||
| PCT/EP2023/070868 WO2024056261A1 (de) | 2022-09-15 | 2023-07-27 | Verfahren zum trainieren eines neuronalen netzes zur detektion eines objekts und verfahren zur detektion eines objekts mittels eines neuronalen netzes |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| EP4587856A1 true EP4587856A1 (de) | 2025-07-23 |
Family
ID=87557990
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| EP23751280.1A Pending EP4587856A1 (de) | 2022-09-15 | 2023-07-27 | Verfahren zum trainieren eines neuronalen netzes zur detektion eines objekts und verfahren zur detektion eines objekts mittels eines neuronalen netzes |
Country Status (5)
| Country | Link |
|---|---|
| US (1) | US20260100029A1 (de) |
| EP (1) | EP4587856A1 (de) |
| CN (1) | CN119836581A (de) |
| DE (1) | DE102023003086A1 (de) |
| WO (1) | WO2024056261A1 (de) |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10430641B2 (en) * | 2017-03-08 | 2019-10-01 | GM Global Technology Operations LLC | Methods and systems for object tracking using bounding boxes |
| DE102018203684A1 (de) | 2018-03-12 | 2019-09-12 | Zf Friedrichshafen Ag | Identifikation von Objekten mittels Radardaten |
| US11899099B2 (en) * | 2018-11-30 | 2024-02-13 | Qualcomm Incorporated | Early fusion of camera and radar frames |
| US11393097B2 (en) * | 2019-01-08 | 2022-07-19 | Qualcomm Incorporated | Using light detection and ranging (LIDAR) to train camera and imaging radar deep learning networks |
| DE102019200141A1 (de) * | 2019-01-08 | 2020-07-09 | Conti Temic Microelectronic Gmbh | Verfahren zum Erfassen von Teilbereichen eines Objekts |
| US10408939B1 (en) * | 2019-01-31 | 2019-09-10 | StradVision, Inc. | Learning method and learning device for integrating image acquired by camera and point-cloud map acquired by radar or LiDAR corresponding to image at each of convolution stages in neural network and testing method and testing device using the same |
| US10776673B2 (en) | 2019-01-31 | 2020-09-15 | StradVision, Inc. | Learning method and learning device for sensor fusion to integrate information acquired by radar capable of distance estimation and information acquired by camera to thereby improve neural network for supporting autonomous driving, and testing method and testing device using the same |
| EP3832341B1 (de) | 2019-11-21 | 2026-01-28 | NVIDIA Corporation | Tiefes neuronales netzwerk zur erkennung von hinderniszuständen mittels radarsensoren in autonomen maschinenanwendungen |
| DE102019219894A1 (de) | 2019-12-17 | 2021-06-17 | Zf Friedrichshafen Ag | Vorrichtung und Verfahren zur Erzeugung von verifizierten Trainingsdaten für ein selbstlernendes System |
| DE112021000135T5 (de) | 2020-06-25 | 2022-06-30 | Nvidia Corporation | Sensorfusion für anwendungen autonomer maschinen durch maschinelles lernen |
-
2023
- 2023-07-27 WO PCT/EP2023/070868 patent/WO2024056261A1/de not_active Ceased
- 2023-07-27 CN CN202380066340.1A patent/CN119836581A/zh active Pending
- 2023-07-27 DE DE102023003086.4A patent/DE102023003086A1/de active Pending
- 2023-07-27 US US19/112,280 patent/US20260100029A1/en active Pending
- 2023-07-27 EP EP23751280.1A patent/EP4587856A1/de active Pending
Also Published As
| Publication number | Publication date |
|---|---|
| WO2024056261A1 (de) | 2024-03-21 |
| DE102023003086A1 (de) | 2024-03-21 |
| CN119836581A (zh) | 2025-04-15 |
| US20260100029A1 (en) | 2026-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3540457A1 (de) | Identifikation von objekten mittels radardaten | |
| EP3695244B1 (de) | Verfahren und vorrichtung zum erzeugen eines inversen sensormodells und verfahren zum erkennen von hindernissen | |
| DE102009014105A1 (de) | Erkennungssystem für ein Fahrzeug | |
| WO2021170321A1 (de) | Verfahren zum detektieren bewegter objekte in einer fahrzeugumgebung und kraftfahrzeug | |
| DE102020127001A1 (de) | Bildbasierte dreidimensionale fahrbahnerkennung | |
| EP4176424B1 (de) | Ampelspurzuordnung aus schwarmdaten | |
| EP3582139A1 (de) | Kamera-basiertes andocken von fahrzeugen mittels künstlicher intelligenz | |
| DE112021004200T5 (de) | Objekterkennungsvorrichtung | |
| DE102019130204A1 (de) | Verfahren und System zum Erstellen dynamischer Karteninformation, die zum Bereitstellen von Umgebungsinformationen geeignet ist | |
| DE102021214760A1 (de) | Verfahren zum Trainieren einer radarbasierten Objektdetektion und Verfahren zur radarbasierten Umfelderfassung | |
| WO2022079162A1 (de) | System und verfahren zum annotieren von automobilradardaten | |
| DE102022211463A1 (de) | Computerimplementiertes Verfahren und Vorrichtung zur Bestimmung einer Klassifikation für ein Objekt | |
| DE102019213927A1 (de) | Grid-basierte Leitpfosten-Klassifizierung | |
| DE102019101127A1 (de) | Verfahren zum Verarbeiten mittels Ultraschall erhaltener Daten | |
| DE102019107650A1 (de) | Objektformerfassung basierend auf zwei Typen von Umgebungssensoren | |
| EP4587856A1 (de) | Verfahren zum trainieren eines neuronalen netzes zur detektion eines objekts und verfahren zur detektion eines objekts mittels eines neuronalen netzes | |
| DE102020118318A1 (de) | Verfahren und Vorrichtung zur Erkennung eines Verkehrsknotenpunktes auf Basis von Trajektoriendaten | |
| DE102017220506B4 (de) | Verfahren und Vorrichtung zum Schätzen eines maritimen Verkehrszustandes | |
| WO2022258444A1 (de) | Verfahren und vorrichtung zum ermitteln einer reichweite eines sensors | |
| DE102023105035A1 (de) | Ein Verfahren zum Bereitstellen von Fehlalarminformationen für Erfassungen, die durch Durchführen mindestens einer Radarmessung erhalten werden | |
| DE102019218932B3 (de) | Ermitteln einer Relativbewegung | |
| DE102006051091B4 (de) | Verfahren zur Objekterkennung von Fahrzeugen mittels Nahbereichsdetektion | |
| DE102020210116A1 (de) | Verfahren zum Betreiben eines Assistenzsystems eines Fahrzeugs | |
| DE102020208544A1 (de) | Hybride Auswertung von Radardaten für die Klassifikation von Objekten | |
| DE102023119911B4 (de) | Verfahren und System zur automatisierten Validierung von Funktionen eines Fahrerassistenzsystems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: UNKNOWN |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
| PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
| 17P | Request for examination filed |
Effective date: 20250415 |
|
| AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC ME MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
| DAV | Request for validation of the european patent (deleted) | ||
| DAX | Request for extension of the european patent (deleted) | ||
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
| 17Q | First examination report despatched |
Effective date: 20260319 |