Image labelling system and method therefor

1 . A system for generating a labelled dataset, the system comprising:

processing means configured to:

receive first data from one or more first sensors wherein the first data comprises one or more frames, each frame comprising an image or a point cloud and further comprising an associated timestamp indicating a time at which the image or point cloud was captured, and wherein the first data comprises data defining an object of interest within a predetermined area;

receive second data from one or more second sensors wherein the second data is associated with the object of interest within the predetermined area, and wherein the second data comprises one or more unique object identifiers and a plurality of timestamps, at least one timestamp of the plurality of timestamps representing a time at which the object of interest entered the coverage of the one or more second sensors, and at least one timestamp of the plurality of timestamps representing a time at which the object of interest left the coverage of the one or more second sensors;

analyse the one or more frames of the first data to identify, based on the second data, the object of interest present in the first data;

label the one or more frames of the first data based on the analysis to generate a labelled dataset;

output the labelled dataset.

2 . The system of claim 1 , wherein at least one of the one or more first sensors is a video camera configured to capture images of an area, and wherein at least one of the one or more second sensors is configured to detect radio frequency signals.

3 . The system of claim 2 , wherein each of the one or more first sensors and each of the one or more second sensors is configured such that the field of view or coverage of the one or more first sensors and the one or more second sensors is aligned.

4 . The system of claim 1 , wherein the first data is associated with two dimensional or three dimensional image data of the predetermined area, and wherein the second data is associated with radio frequency data.

5 . The system of claim 1 , wherein the first data comprises a single frame of video data and the timestamp corresponds to the time at which the single frame of video data was detected and/or wherein the first data comprises a single instance of point cloud depth data and the timestamp corresponds to the time at which the single instance of point cloud depth data was detected.

6 . The system of claim 1 , wherein the one or more unique identifiers are internet protocol, IP, addresses.

7 . The system of claim 1 , wherein the processing means is further configured to receive positioning data indicating the position of the object of interest, and analyse the one or more frames of the first data to identify, based on the second data and the positioning data, the object of interest present in the one or more frames of the first data.

8 . A detection system for detecting one or more objects of interest, the detection system comprising:

an input module configured to receive first sensor data comprising one or more frames;

a machine learning model trained using a training system comprising a processing means configured to:

receive data labelled using the system of claim 1 ; and

train the machine learning model using the received data; and

a processing means configured to:

determine, using the machine learning model, whether any objects of interest are present in each frame of the one or more frames of the first sensor data.

9 . The detection system of claim 8 , wherein the detection system is further configured to track one or more vehicles within the field of view of the first sensor data, based on the determination.

10 . The detection system of claim 9 , wherein the detection system is further configured to determine a likelihood of a collision between the one or more tracked vehicles.

11 . The detection system of claim 8 , wherein the detection system is further configured to detect when an object of interest enters the predetermined area and generate an entry timestamp corresponding to the time at which the object of interest enters the predetermined area.

12 . The detection system of claim 8 , wherein the detection system is further configured to detect when the object of interest exits the predetermined area and generate an exit timestamp corresponding to the time at which the object of interest leaves the predetermined area.

13 . The system of claim 1 , wherein the system is further configured to determine a period of time during which the object of interest is within the predetermined area.

14 . The system of claim 13 , wherein the system is further configured to generate the labelled dataset only when the period of time is greater than a predetermined threshold.

15 . The system of claim 1 , further comprising a module which is configured to adjust the range of the one or more second sensors in response to a range adjust command.

16 . The system of claim 1 , wherein the system is further configured to determine a size of a further object or a length or width of a further object and preferably wherein the size, length or width is determined by comparing a detected identifier associated with the further object with a look-up table of identifiers and associated sizes, lengths or widths of the further object.

17 . The system of claim 16 , further comprising a module which is configured to adjust the range of the one or more second sensors based on the determined size or length or width of the further object.

18 . The system of claim 1 , wherein the system is further configured to determine one or more sub sectors within the predetermined area based on triangulation of radio frequency signals from a plurality of signal detectors, wherein the labelled data set is only generated for objects which are not within one or more of the subsectors.

19 . A method for generating a labelled dataset, the method comprising:

receiving first data from one or more first sensors wherein the first data comprises one or more frames, each frame comprising an image or a point cloud and further comprising an associated timestamp indicating a time at which the image or point cloud was captured, and wherein the first data comprises data defining an object of interest within a predetermined area;

receiving second data from one or more second sensors wherein the second data is associated with the object of interest within the predetermined area, and wherein the second data comprises one or more unique object identifiers and one or more timestamps representing times at which the object of interest entered or left the coverage of the one or more second sensors;

analysing the one or more frames of the first data to identify, based on the second data, the object of interest present in the first data; and

labelling the one or more frames of the first data based on the analysis to generate a labelled dataset.