Method for object detection and optical detection system

Abstract

The present disclosure relates to a method for object detection using first and second optical data (21, 22) of a detection area (32), the first optical data (21) comprising active image data, the second optical data (22) comprising passive image data, the method comprising receiving the first and second optical data (21, 22) from an optical reception sensor (12), the optical reception sensor (12) comprising a plurality of pixels, the pixels being configured to provide the first and second optical data (21, 22) of the detection area (32), performing object detection in the detection area (32) using the first and second optical data (21, 22). The present disclosure further relates to an optical detection system (10) and a vehicle (30) comprising the optical detection system (10).

Description

[0001] METHOD FOR OBJECT DETECTION AND OPTICAL DETECTION SYSTEM

[0002] Field

[0003] The present disclosure relates to a method for object detection using first and second optical data, an optical detection system and a vehicle comprising such an optical detection system.

[0004] Background

[0005] Modern vehicles like cars, vans, trucks, motorcycles, etc. may comprise sensor systems, whose data are used for driver information and / or are used by driver assistance systems.

[0006] Sensor systems are constantly being developed for various functions, e. g. for the acquisition of environmental information in the near and far range of vehicles, such as passenger cars or commercial vehicles. Based on the acquired data, a model of the vehicle environment can be generated and a reaction to changes in this vehicle environment is possible. Sensor systems can also serve as sensors for driver assistance systems, in particular assistance systems for autonomous or semi-auton- omous vehicle control. They can for example be used to detect obstacles and / or other road users in the front, rear or blind spot areas of a vehicle. Sensor systems can be based on different sensor principles, such as radar, ultrasound, optics.

[0007] Optical detection systems comprise passive optical detection systems like e. g. a camera and active optical detection systems like e. g. a lidar system (lidar: light detection and ranging). A lidar system comprises an optical transmission device and an optical reception device. The transmission device emits an optical signal, which can be continuous or pulsed. In addition, the optical signal may be modulated. For example, electromagnetic waves in the form of laser beams in the ultraviolet, visual or infrared range may be used as optical signals in a lidar system. The light is received by the optical reception device after reflection from an object in a detection area of the lidar system. The optical signal can for example be evaluated according to a time-of-flight method and the spatial position and distance of the object on which the reflection occurred can be determined. In addition, it may be possible to determine a relative velocity. Reflection or reflected light is understood to mean any light that is reflected back and should also include, in particular, light that is reflected back by scattering or absorption emission.

[0008] An optical reception sensor of an optical detection systems may comprise several receiving elements, so-called pixels. A pixel of the optical reception sensor generates an electrical signal in response to the reception of an optical signal. The pixels may be set up to receive light from different receiving angles.

[0009] In US2016 / 0054434 a dual-mode imaging device is described. The imaging device comprises a pixelated array of semiconductor detector elements, in which each detecting element is electrically connected to an integrated circuit, the integrated circuit of each of the pixels comprising a passive signal path and a transient signal path. The passive path provides consecutive frame or scene imaging and the transient path detects the transient electromagnetic events such as laser pulses. The passive path and the transient path operate simultaneously.

[0010] In US2019 / 0353791A1 an electronic device is described which is configured to receive first image data of a passive optical sensor and receive second data of an active sensor. The data of the active sensor is used for augmenting the image data of the passive optical sensor.

[0011] Summary

[0012] A method for object detection uses first and second optical data of a detection area. The first optical data comprises active image data, the second optical data comprises passive image data. The method comprises:

[0013] • Receiving the first and second optical data from an optical reception sensor, the optical reception sensor comprising a plurality of pixels, the pixels being configured to provide the first and second optical data of the detection area.

[0014] • Performing object detection in the detection area using the first and second optical data.

[0015] The use of the first and second optical data for performing the object detection allows for a self-validation of the object detection. The use of the first and second optical data provided by the same optical reception sensor allows to perform the object detection including the self-validation in a very short timeframe, fore example in real-time, e. g. within nanoseconds. The optical reception sensor comprises pixels that comprise photosensitive devices that capture and convert light into an electrical signal that can be further processed, e. g. measured, analyzed, and interpreted. The photosensitive devices may comprise photodiodes, charge-coupled devices (CCD), CMOS sensors, photomultiplier tubes, Single Photon Avalanche Diodes (SPADs), or similar.

[0016] The optical reception sensor provides both the first and second optical data. It is configured to provide both the active image data and the passive image data of the detection area. A respective pixel is configured to output the active image data and the passive image data. The detection area is an area in the environment of the optical reception sensor from where the optical signals are captured.

[0017] The optical reception sensor may be comprised in an optical detection system. The optical detection system uses optical signals to detect and / or measure objects and / or conditions in its detection area. The detection area is an area in the environment of the optical detection system which is covered by the optical detection system. The optical detection system is a combined active and passive optical detection system, where the optical reception sensor comprised in the optical detection system is configured to detect both first and second optical signals and to output active and passive image data using the first and second optical signals.

[0018] Active image data is obtained using optical signals emitted by an optical source of the optical detection system. Such emitted optical signals are for example generated by lasers or LED comprised in the optical source. The emitted optical signal is then reflected back in the detection area and analyzed. Active image data is obtained by actively illuminating the detection area, capturing and analyzing the reflected light. Examples of active optical detection systems comprise lidar systems.

[0019] Passive image data is obtained relying on ambient light or thermal emissions from e. g. objects in the detection area. Passive image data is obtained by capturing and analyzing light or radiation that naturally emanates from or is reflected in the detection area.

[0020] In some embodiments the first optical data comprising the active image data may be captured when the reflected light from an illumination period is expected back. The first optical data may in particular comprise 3D image data.

[0021] 3D image data is sometimes referred to as point cloud data. The points of the point cloud comprise information about the spatial location and the depth of the points of reflection in the environment. The points of reflection are the points where the reflection of the optical signal emitted and received by the optical detection system occurred. The point cloud can be understood as a set of points, where each point contains corresponding coordinates in a particular three-dimensional coordinate system. In the case of a three-dimensional point cloud, for example, the three- dimensional coordinates can be determined by direction of incidence of a light beam reflected at a point of reflection and the corresponding propagation time or radial distance, also referred to as depth, measured for that particular point. The direction of incidence may for example be given by the horizontal angle and the elevation angle. In addition to the spatial information, e.g. the three-dimensional coordinates that comprise the distance or depth to the point, the point cloud can also store additional information or measurement data for the individual points, such as the pulse width of the respective received optical signal.

[0022] The three-dimensional coordinate system of the point cloud may be a three-dimensional polar coordinate system. However, the information can also be given in Cartesian coordinates for each of the points. A conversion between the coordinate systems is possible. In addition to the 3D image data, the point cloud may also comprise 2D image data.

[0023] The second optical data comprising the passive image data may be captured by collecting light over periods of time and / or during intervals distinct from the intervals when active image data is captured. The second optical data may comprise 2D image data, for example grayscale image data.

[0024] A grayscale image, sometimes called gray level image, is a type of digital image that contains only shades of gray, varying from black at the weakest intensity to white at the strongest. Unlike color images that are represented using multiple channels, e. g. red, green, and blue, grayscale images have a single channel, where each pixel represents the intensity of light at that point, without color information. In the context of digital imaging, each pixel in a grayscale image has a value representing the brightness of that pixel. The use of the grayscale image data in image processing has the advantage that the amount of data to be processed is reduced thus simplifying the required processing.

[0025] The point cloud may comprise the first optical data, comprising the active image data, and the second optical data, comprising the passive image data. The point cloud may be generated by the optical detection system. The optical detection system may be comprised in a vehicle. The point cloud data may be further processed within the vehicle, e. g. by a computing unit. Such further processing may for example comprise further processing of the point cloud and / or its use by autonomous or partly autonomous driving functions.

[0026] The object detection comprises identifying and localizing objects within the first and second optical data. The objects may be output together with the second optical data in a graphical manner visualizing the location, shape etc. of the object on the passive image of the environment of the detection area. The object identification may comprise a feature extraction step. In this step relevant features that can help in identifying objects are extracted from the optical data. Such extracted features may include local geometric features like normals and / or curvatures.

[0027] In an embodiment of the method the detected objects may be output as a list - optionally together with the second optical data. The list of detected objects may be further used by other algorithms to perform tasks like e. g. autonomous or semi-autonomous driving functions.

[0028] In an embodiment of the method, the object detection further comprises performing clustering, edge detection and / or segmentation on the first optical data and separately on the second optical data. The segmentation and / or clustering of the optical data into clusters and / or segments that potentially represent individual objects or parts of objects is done separately on the first and second optical data. The output of the respective separate steps can then be used to check and / or improve and / or correct the output of the respective other step.

[0029] In an embodiment of the method, the object detection using the first optical data comprises a classification step using artificial intelligence. The classification may comprise a feature extraction step, where significant and informative attributes of first optical data are identified and extracted. These features could be related to the shape, texture, or geometric properties of objects within first optical data. In some approaches, deep learning models, may be used to automatically learn and extract these features directly from the data, without the need for manual feature design. With the features extracted, a classification model, e. g. neural network, is trained in the next step. Once trained, the model can perform the classification.

[0030] In an embodiment of the method, the object detection using the second optical data comprises image signal processing. Image signal processing (ISP) may involve a series of steps to convert the second optical data captured by the optical reception sensor. Examples may include filtering, demosaicing, interpolation, color correction, noise reduction, sharpening, lens distortion correction, chromatic aberration correction, color space correction and / or compression. These steps can be implemented in various orders and combinations.

[0031] In embodiments of the method, the object detection using the first optical data comprises generating at least one list of candidate objects using the first optical data. The output of the object detection step or algorithm may thus be a list of candidate objects that have been identified in the first optical data comprising active image data.

[0032] In embodiments of the method, the at least one list of candidate objects comprises a list of dynamic candidate objects and a list of static candidate objects There may be a list of static candidate objects generated using the first optical data and there may be a list of dynamic candidate objects generated using the first optical data. Dynamic objects are objects that are moving. The object detection used to detect the objects in the first optical data may additionally be able to detect if the object is static or dynamic.

[0033] In embodiments of the method, the object detection using the second optical data comprises at least one further list of candidate objects using the second optical data. The output of the object detection step or algorithm may thus be a further list of candidate objects that have been identified in the second optical data comprising passive image data.

[0034] In embodiments of the method, the at least one further list of candidate objects comprises a further list of dynamic candidate objects and a further list of static candidate objects. There may be a further list of static candidate objects generated using the second optical data and there may be a further list of dynamic candidate objects generated using the second optical data. The object detection used to detect the objects in the second optical data may additionally be able to detect if the object is static or dynamic.

[0035] From the at least one list of candidate objects and the at least one further list of candidate objects, the list of objects may be generated. The list of objects comprises the objects that have been identified by the described method for object detection. The list of objects may then be output.

[0036] In an embodiment of the method, a confidence measure is calculated using the at least one list of candidate objects and the at least one further list of candidate objects. The confidence measure is then output together with the object list. The use of the object list in further processing steps may thus be guided by the confidence measure. If the confidence measure is too low, this may for example be taken into account for the further processing steps, in particular if safety is at stake. For example, the confidence measure may influence the trajectory planning and / or emergency braking handling by a driving function in the vehicle.

[0037] An optical detection system comprises the optical reception sensor. The optical reception sensor comprises a plurality of pixels, the pixels being configured to provide the first and the second optical data of the detection area. The first optical data comprises active image data, the second optical data comprises passive image data. The optical detection system further comprises a processor configured to perform object detection in the detection area using the first and second optical data.

[0038] Using the first and second optical data provided by the same optical reception sensors allows to improve the quality of the object detection of the optical detection system. The first and second optical data relates to the same detection area covered by the optical reception sensor. The processing speed can be fast, as the processing may be done where the first and second optical data is generated by the optical reception sensor.

[0039] The optical detection system may further comprise an optical source used for obtaining the active image data, wherein the optical reception sensor and the optical source are comprised in a real-time environment. The real-time environment is configured such that it allows for the real-time processing within tight time constraints as required by real-time applications. By placing the optical reception sensor and the optical source within the same real-time processing environment, the object detection may be performed at such high speed that it may be applied to functions with tight time-constraints and the need for highly accurate and timely data.

[0040] In embodiments of the optical detection system the optical reception sensor and the optical source are arranged on a single chip. This may further ensure quick processing under tight time constraints.

[0041] In an embodiment of the optical detection system, the processor is configured to perform the steps of the described method for object detection. The optical detection system may be comprised in a vehicle. The output of the optical detection system, e. g. the generated optical data and / or the list of detected objects may be further processed by e. g. a computing unit within the vehicle.

[0042] Brief description of the figures

[0043] Embodiments will now be described with reference to the attached drawing figures by way of example only. Like reference numerals are used to refer to like elements throughout. The illustrated structures and devices are not necessarily drawn to scale.

[0044] Fig. 1 schematically illustrates a method for object detection using first and second optical data.

[0045] Fig. 2 schematically illustrates a vehicle with an optical detection system.

[0046] Fig. 3 schematically illustrates lists of candidate objects.

[0047] Fig. 4 schematically illustrates further lists of candidate objects.

[0048] Fig. 5 schematically illustrates a list of objects.

[0049] Fig. 6 schematically illustrates an optical detection system.

[0050] Detailed Description

[0051] Figure 1 schematically illustrates a method for object detection using first optical data 21 and second optical data 22 of a detection area 32. The first optical data 21 comprises active image data and the second optical data 22 comprises passive image data.

[0052] In 100, the first and second optical data 21, 22 is received from an optical reception sensor 12. The optical reception sensor 12 comprises a plurality of pixels, the pixels being configured to provide the first and second optical data 21, 22 of the detection area (32).

[0053] In 102, the object detection in the detection area 32 is performed using the first optical data 21.

[0054] In 104, the object detection in the detection area 32 is performed using the second optical data 22.

[0055] In 106, the information obtained in 102 and 104 is fused.

[0056] In 108 the information obtained in 106 is validated. In 110 the information obtained in 108 is output.

[0057] Figure 2 schematically illustrates a vehicle 30, for example a passenger car. The vehicle 30 comprises an optical detection system 10. The optical detection system 10 is arranged in a front area of the vehicle 30. It comprises a processor 20 and a real-time environment 16, e. g. a real-time chip. The real-time environment 16 comprises an optical source 14 and an optical reception sensor 12. The optical source 14 is configured to transmit optical signals 36 into the detection area 32. After reflection in the detection area 36, e. g. on an object O, the optical signals 36 are received by the optical reception sensor 12.

[0058] The optical detection system 10 may optionally comprise an optical deflection device that may be arranged such that it deflects the optical signal 36 sent by the optical source 14 and to be received by the optical reception sensor 12. The optical deflection device may be controlled such that the optical signal L performs a scanning movement over the detection area 32. The optical deflection device may for example comprise a rotating mirror that performs a rotating movement to deflect the optical signal 36 such that the scanning movement is performed by the optical signal 36.

[0059] The optical detection system 10, is for example a lidar system which has the capability to capture the second optical data comprising the passive image data 22 and first optical data comprising the active image data 21 with one single optical reception sensor 12 located on the real-time environment 16, e. g. on one single real-time chip.

[0060] The second optical data 22 comprising the passive image data is captured by the optical reception sensor 12 by capturing ambient light similar to a camera.

[0061] The first optical data 21 comprising the active image data is generated by the optical detection system 10 by sending out the optical signal 36 using the optical source 14 and capturing the optical signal 36 reflected in the detection area 32.

[0062] The processor 20 may control the transmission and reception process of the optical signal 36 and the reception of the second optical data 22 comprising the passive image data.

[0063] In the processor 20, the transmitted and received optical signals 36 may be evaluated e. g. using time-of-flight measurements in order to obtain the first optical data 21. The first optical data 21 contains the collection of measurement data for the various points of reflection of the optical signal 36 in the coverage area 32. For example, each pulse of light of the optical signal 36 may provide a data point.

[0064] The second optical data 22 contains the collection of received passive image data of the detection area 32. This data may be used to improve and verify the first optical data 21.

[0065] Object detection in the detection area 32 is then performed by the processor 20 using the first and second optical data 21, 22. The data generated in this way can be transferred from the processor 20 to a computing unit 34 of vehicle 30.

[0066] The optical detection system 10 may for example be placed or integrated at the front of the vehicle 30. Thus, an area in front of the vehicle 30 in the direction of travel can be monitored by the optical detection system 10. Detection systems 10 may be placed on other parts of the vehicle 30, e. g. for surround-view functions such as at the sides and / or rear of the vehicle 30. It is also possible to arrange detections systems and / or other sensors like radar, ultrasound etc. on the vehicle 30, also in corner areas of the vehicle 30. Sensor fusion may then be performed using the sensor data of the various sensors.

[0067] The optical detection system 10 may be used to detect features of the environment. The features may be stationary or moving objects in the environment. Features may comprise vehicles, persons, animals, plants, obstacles, roadway unevenness, in particular potholes or stones, roadway boundaries, traffic signs, open spaces, in particular parking spaces, precipitation, or the like.

[0068] The optical detection system 10 may provide an accurate, dense and reliable point cloud of the detection area 32. The point cloud may show the contour of objects and / or other features. The data has been verified by fusing the first optical data with the second optical data, thus making the data more reliable and accurate. Also, a confidence measure may be created to give an indication of the quality of the data.

[0069] The computing unit 34 may be designed as a central vehicle computer of the vehicle 30, in which data from several detection systems of vehicle 30 may be received, evaluated and / or further processed. The computing unit 34 may be used, for example, to implement autonomous or semi-autonomous driving functions.

[0070] Figure 3 schematically illustrates lists of candidate objects LS-21, LD-21 which were created using the first optical data 21. A first list of candidate objects LS-21 comprises static candidate objects SOA-1, SOA-2, ... , SOA-nl that were detected in the first optical data 21. A second list of candidate objects LD-21 comprises dynamic candidate objects DOA-1, DOA-2, ... , DOA-ml that were detected in the first optical data 21.

[0071] The lists of candidate objects LS-21, LD-21 are generated by the processor 20 by processing the first optical data 21 using an artificial intelligence based classification algorithm, e. g. a trained neural network. The classification algorithm allows to detect the dynamic and static objects DO-1, DO-2, ... , DO-m, SO-1, SO-2, ... , SO-n in the detection area 32.

[0072] Figure 4 schematically illustrates further lists of candidate objects LS-22, LD-22 which were created using the second optical data 22.

[0073] A first further list of candidate objects LS-22 comprises static candidate objects SOP-1, SOP-2, ... , SOP-n2 that were detected in the second optical data 22. A second further list of candidate objects LD-22 comprises dynamic candidate objects DOP-1, DOP-2, ... , D0P-m2 that were detected in the second optical data 22.

[0074] The further lists of candidate objects LS-22, LD-22 are generated by the processor 20 by processing the second optical data 22 using image signal processing. The image signal processing allows to detect the dynamic and static objects DO-1, DO- 2, ... , DO-m, SO-1, SO-2, ... , SO-n in the detection area 32.

[0075] Figure 5 schematically illustrates the list of objects OL in the detection area 32. It comprises a static object list SOL and a dynamic object list DOL. The dynamic object list DOL comprises the dynamic objects DO-1, DO-2, ... , DO-m, in the detection area 32. The static object list SOL comprises the static objects SO-1, SO- 2, ... , SO-n in the detection area 32.

[0076] For each object, it is identified if it has been detected in the first optical data 21 and / or in the second optical data 22. An identified object is marked with a "+". An unidentified object is marked with a "- ". By using this list OL for comparison, the object detection can be improved by counter checking. For example, only objects detected in both the first and second optical data 21, 22 could be regarded as a verified object. Also, a confidence measure could be created from the dynamic object list DOL and the static object list SOL.

[0077] During the real-time operation of the optical source 14 and the optical reception sensor 12 it needs to be decided in real-time about the reliability of the object detection. The use of the first and second optical data 21, 22 can support in assessing this question and allow to assess the reliability and confidence of the information provided by the optical detection system 10. This supports to assess the reliability of the object information, by allowing for qualifying and validating the generated object lists LS-21, LD-21, LS-22, LD-22. This qualification and validation may be done internally to the processor 20.

[0078] Figure 6 schematically illustrates an optical detection system 10 performing the described method for object detection comprising self-validation.

[0079] The optical reception sensor 12 is configured to perform both active and passive sensing in real time. The optical reception sensor 12 is comprised in the real-time environment 16. The real-time environment also comprises the optical source 14. The real-time environment 16 may be realized on a single chip and may allow the processor 20 to perform the object detection including the self-validation within e. g. nanoseconds.

[0080] The optical reception sensor 16 outputs the detected first and second optical data 21, 22 to the processor 20 as point cloud data 18. The point cloud data 18 comprises the first optical data 21 and the second optical data 22.

[0081] In the next steps, the first optical data 21 comprising active image data, e. g. lidar data, is processed separately from the second optical data 22 comprising passive image data, e. g. grayscale image.

[0082] In step 102, the object detection in the detection area 32 is performed using the first optical data 21.

[0083] In 102.1 preprocessing is performed on the first optical data 21. Preprocessing comprises data cleaning to remove noise and outliers to improve the quality of the data for object detection. Preprocessing further comprises down sampling to reduce the size of the data set to make the processing more manageable, using techniques like e. g. voxel grid down sampling.

[0084] In 102.2 clustering and feature extraction like edge detection is performed on the output of 102.1. Clustering comprises grouping points according to similarity and feature extraction transforms the point data information into feature information. Features can be edges, corners, textures or similar. In 102.3 segmentation is performed on the output of 102.2. Segmentation comprises partitioning the image data into multiple segments that potentially represent objects. The output of 102.3 are the candidate object lists LS-21, LD-21.

[0085] In 104, the object detection in the detection area 32 is performed using the second optical data 22.

[0086] In 104.1 image signal processing is performed on the second optical data 22.

[0087] In 104.2 clustering and feature extraction like edge detection is performed on the output of 104.1. Clustering comprises grouping points according to similarity and feature extraction transforms the point data information into feature information. Features can be edges, corners, textures or similar.

[0088] In 104.3 segmentation is performed on the output of 104.2. Segmentation comprises partitioning the image data into multiple segments that potentially represent objects. The output of 104.3 are the candidate object lists LS-22, LD-22.

[0089] In 106, the information obtained in 102 and 104 is fused. The fusion step 106 combines the lists LS-21, LD-21, LS-22, LD-22 described with respect to Figures 3 and 4. In 108 the information obtained in 106 is validated. The validation step 108 has been described with respect to Figures 5. The output of 108 is the list of detected and validated objects OL. Together with the list of detected and validated objects OL the second optical data 22 may be output.

[0090] The described method and optical detection system 10 allow to enhance the reliability of the detected object list OL. Furthermore, the performance may be improved by reducing false positives. Further, it is possible to assess the object detection qualitatively and quantitatively, e. g. by determining a confidence measure, like e. g. a positive predictive value (PPV) or similar.

[0091] In particular ghost objects may be identified using the described method. Ghost objects are objects that appear in the image data but do not exist in reality.

[0092] The quality of the object detection may be further improved by applying a pixel- to-pixel quality comparison. For example, the same object should have the same pixel size in both the first and second optical data 21, 22. If an object O has for example a size of 24 x 30 pixels in the passive image, in active image the object O should also be 24 x 30 pixels in size. Any artefacts surrounding this pixel area of the real size can then be identified as noise. It can then be filtered, and the performance and quality of the object detection can be improved.

Claims

CLAIMS1. Method for object detection using first and second optical data (21, 22) of a detection area (32), the first optical data (21) comprising active image data, the second optical data (22) comprising passive image data, the method comprising receiving the first and second optical data (21, 22) from an optical reception sensor (12), the optical reception sensor (12) comprising a plurality of pixels, the pixels being configured to provide the first and second optical data (21, 22) of the detection area (32), performing object detection in the detection area (32) using the first and second optical data (21, 22).

2. Method according to claiml, further comprising outputting a list of detected objects (OL) and the second optical data (22).

3. Method according to claim 1 or 2, the object detection further comprising performing clustering, edge detection and / or segmentation on the first optical data (21) and separately on the second optical data (22).

4. Method according to one of the preceding claims, wherein the object detection using the first optical data (21) comprises a classification step using artificial intelligence.

5. Method according to one of the preceding claims, wherein the object detection using the second optical data (22) comprises image signal processing.

6. Method according to one of the preceding claims, wherein the object detection using the first optical data (21) comprises generating at least one list of candidate objects (LS-21, LD-21) using the first optical data (21) and / or the object detection using the second optical data (22) comprises at least one further list of candidate objects (LS-22, LD-22) using the second optical data (22).

7. Method according to claim 6, wherein the at least one list of candidate objects (LS-21, LD-21) comprises a list of dynamic candidate objects (LD-21) and a list of static candidate objects (LS-21) and / or the at least one further list of candidate objects (LS-22, LD-22) comprises a further list of dynamic candidate objects (LD-22) and a further list of static candidate objects (LS-22).

8. Method according to claim 6 or 7, wherein the list of objects (OL) is generated using the at least one list of candidate objects (LS-21, LD-21) and the at least one further list of candidate objects (LS-22, LD-22).

9. Method according to one of claims 6 to 8, wherein a confidence measure is calculated and output using the at least one list of candidate objects (LS-21, LD-21) and the at least one further list of candidate objects (LS- 22, LD-22).

10. Optical detection system (10) comprising an optical reception sensor (12), the optical reception sensor (12) comprising a plurality of pixels, the pixels being configured to provide first and second optical data (21, 22) of a detection area (32), the first optical data (21) comprising active image data, the second optical data (22) comprising passive image data, the optical detection system (10) further comprising a processor (20) configured to perform object detection in the detection area (32) using the first and second optical data (21, 22).

11. Optical detection system according to claim 10, further comprising an optical source (14) used for obtaining the active image data, wherein the optical reception sensor (12) and the optical source (14) are comprised in a real-time environment.

12. Optical detection system according to claim 10 or 11, the optical reception sensor (12) and the optical source (14) being arranged on a single chip.

13. Optical detection system according to one of claims 10 to 12, wherein the processor (20) is configured to perform the steps of the method according to one of claims 1 to 9.

14. Vehicle (30) comprising the optical detection system (10) according to one of claims 10 to 13.

Images