Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020142036 - INDOOR DRONE NAVIGATION SUPPORT SYSTEM BASED ON MACHINE LEARNING

Note: Text based on automatic Optical Character Recognition processes. Please use the PDF version for legal matters

[ EN ]

CLAIMS

1. A system for positioning an unmanned aerial vehicle (100) in a setting (400) comprising at least one first camera (131) at the unmanned aerial vehicle (100) arranged to capture at least one image comprising at least one of the unique markers (300) arranged in an order along at least one first axis (401) in the said setting (400) in which the unmanned aerial vehicle (100) flies,

the unmanned aerial vehicle (100) comprises a processor unit (110)

- accessing a first marker data set (1421) stored in a memory unit (140) comprising at least one marker image for each said marker (300),

- applying a first classification algorithm to the image to generate a first prediction in which the marker in the image belongs to which of the images in the first marker dataset (1421) is predicted,

- generating the second prediction in which the marker in the image belongs to which of the marker images in the first marker dataset (1421) is predicted by applying the second classification algorithm, the type of which is different from that of the said first classification algorithm, to the image,

- generating the third prediction in which the marker in the image belongs to which of the images in the first marker dataset (1421) is predicted by applying the third classification algorithm, the type of which is different from those of the said first classification algorithm and the second classification algorithm, to the image,

- ensuring that the marker in the image is matched to the marker image with the highest number of predictions as a result of the said first prediction, the said second prediction, and the said third prediction,

- being configured to access a map stored in the said memory unit (140) to determine the position of the matched marker on the first axis (401) of the map covering the original markers and their position in the setting (400) where the unmanned aerial vehicle flies.

2. A system for positioning an unmanned aerial vehicle in accordance with Claim 1 characterized in that unmanned aerial vehicle (100) comprises at least one second camera (132) arranged to capture at least one image comprising at least one of the unique markers arranged in an order along a second axis (402) perpendicular to the first axis (401),

the processor unit (110)

- accessing the second marker data set (1422) stored in the memory unit (140) comprising at least one marker image for each said marker,

- generating the first prediction in which the marker in the image belongs to which of the marker images in the second marker dataset (1422) is predicted by applying the first classification algorithm to the image,

- generating the second prediction in which the marker in the image belongs to which of the marker images in the second marker dataset (1422) is predicted by applying the second classification algorithm, the type of which is different from that of the said first classification algorithm, to the image,

- generating the third prediction in which the marker in the image belongs to which of the images in the second marker dataset (1422) is predicted by applying the third classification algorithm, the type of which is different from those of the said first classification algorithm and the second classification algorithm, to the image,

- ensuring that the marker in the image is matched to the marker image with the highest number of predictions as a result of the said first prediction, the said second prediction, and the said third prediction,

- being configured to access the map to determine the position of the matched marker on the second axis (402) on the map.

3. A system for positioning an unmanned aerial vehicle in accordance with Claim 2 characterized in that the unmanned aerial vehicle (100) comprises at least one third camera (133) arranged to capture at least one image comprising at least one of the unique markers arranged in an order along a third axis (403) perpendicular to the first axis (401) and the second axis (402),

the processor unit (110)

- accessing the third marker dataset (1423) stored in the memory unit (140) comprising at least one marker image for each said marker,

- generating the first prediction in which the marker in the image belongs to which of the marker images in the library is predicted by applying the first classification algorithm to the image,

- generating the second prediction in which the marker in the image belongs to which of the marker images in the third marker dataset (1423) is predicted by applying the second classification algorithm, the type of which is different from that of the said first classification algorithm, to the image,

- generating the third prediction in which the marker in the image belongs to which of the images in the third marker dataset (1423) is predicted by applying the third classification algorithm, the type of which is different from those of the said first classification algorithm and the second classification algorithm, to the image,

- ensuring that the marker in the image is matched to the marker image with the highest number of predictions as a result of the said first prediction, the said second prediction, and said third prediction,

- being configured to access the map to determine the position of the matched marker on the third axis (403) on the map.

4. A system for positioning an unmanned aerial vehicle in accordance with Claim 1 characterized in that it comprises a user terminal (200), a communication unit (150) which enables unmanned aerial vehicle's (100) wireless data communication with the processor unit (110) and the user terminal (200), the processor unit (110) being arranged to send the location information to the user terminal (200).

5. A system for positioning an unmanned aerial vehicle in accordance with Claim 1 characterized in that the processor unit (1 10) is arranged to control a flight control unit (120) of the unmanned aerial vehicle (100).

6. A system for positioning an unmanned aerial vehicle in accordance with Claim 1 characterized in that said markers comprise a QR code.

7. A system for positioning an unmanned aerial vehicle in accordance with Claim 1 characterized in that each of said markers comprises a unique pattern.

8. A system for positioning an unmanned aerial vehicle in accordance with Claim 3 characterized in that the markers in the first axis (401) are provided in a first color, the markers in the second axis (402) are provided in a second color different from said first color, and the markers in the third axis (403) are provided in a third color different from the first color and the second color.

9. A system for positioning an unmanned aerial vehicle in accordance with Claim 3 characterized in that the first color, the second color, and the third color are selected out of red, blue, and green.

10. A system for positioning an unmanned aerial vehicle in accordance with Claim 1 characterized in that the first classification algorithm is the Random Forest Algorithm, the second classification algorithm is the K-nearest Neighbors Algorithm, and the third classification algorithm is the Logistic Regression Algorithm.

1 1. A system for positioning an unmanned aerial vehicle in accordance with Claim 1 characterized in that a plurality of marker images is provided for each marker and the said marker images are obtained by subjecting the original marker image to at least one distortion process.

12. A system for positioning an unmanned aerial vehicle in accordance with Claim 1 1 characterized in that the marker images in the marker dataset (142) are compressed.

13. A system for positioning an unmanned aerial vehicle in accordance with Claim 12 characterized in that the marker images in the marker dataset (142) are compressed using the Principal Component Analysis (PCA) method.

14. A system for positioning an unmanned aerial vehicle in accordance with any of the Claims 1 , 2, and 3 characterized in that an Ensemble Voting Algorithm is used in the step of“ensuring the marker in the image to match the marker image having the highest number of predictions in the said first prediction, the said second prediction, and said third prediction".

15. A computer-based method for positioning an unmanned aerial vehicle (100) in a setting (400) in which the unmanned aerial vehicle (100) comprises the following steps performed by the processor unit (1 10):

- receiving at least one image which is captured by at least one first camera (131) of the unmanned aerial vehicle and which comprises at least one of the original markers arranged in an order along at least one first axis (401) in the said setting (400) where the unmanned aerial vehicle (100) flies, as input,

- accessing a first marker data set (1421) stored in a memory unit (140) comprising at least one marker image for each said marker,

- generating the first prediction in which the marker in the image belongs to which of the marker images in the first marker dataset (1421) is predicted by applying the first classification algorithm to the image,

- generating the second prediction in which the marker in the image belongs to which of the marker images in the first marker dataset (1421) is predicted by applying the second classification algorithm, the type of which is different from that of the said first classification algorithm, to the image,

- generating the third prediction in which the marker in the image belongs to which of the images in the first marker dataset (1421) is predicted by applying the third classification algorithm, the type of which is different from those of the said first classification algorithm and the second classification algorithm, to the image,

- ensuring that the marker in the image is matched to the marker image with the highest number of predictions as a result of the said first prediction, the said second prediction, and the said third prediction,

- accessing a setting map (143) stored in the said memory unit (140), comprising the original markers and their positions in the setting (400) in which the unmanned aerial vehicle flies, determining the position of the matched marker on the first axis (401) in the setting map (143).

16. A method for positioning an unmanned aerial vehicle in accordance with Claims 15 characterized with the following features:

- receiving at least one image which is captured by at least one second camera (132) of the unmanned aerial vehicle (100) and which comprises at least one of the original markers arranged in an order along at least one second axis (402) in the said setting (400) where the unmanned aerial vehicle (100) flies, as input,

- accessing the second marker data set (1422) stored in the memory unit (140) comprising at least one marker image for each said marker,

- generating the first prediction in which the marker in the image belongs to which of the marker images in the second marker dataset (1422) is predicted by applying the first classification algorithm to the image,

- generating the second prediction in which the marker in the image belongs to which of the marker images in the second marker dataset (1422) is predicted by applying the second classification algorithm, the type of which is different from that of the said first classification algorithm, to the image,

- generating the third prediction in which the marker in the image belongs to which of the marker images in the second marker dataset (1422) is predicted by applying the third classification algorithm, the type of which is different from those of the said first classification algorithm and the second classification algorithm, to the image,

- ensuring that the marker in the image is matched to the marker image with the highest number of predictions as a result of the said first prediction, the said second prediction, and the said third prediction,

- accessing the setting map (143) and being configured to determine the position of the matched marker on the second axis (402) on the map.

17. A method for positioning an unmanned aerial vehicle in accordance with Claims 16 characterized with the following features:

- receiving at least one image which is captured by at least one third camera (133) of the unmanned aerial vehicle (100) and comprising at least one of the original markers arranged in an order along at least one third axis (403) in the said setting (400) where the unmanned aerial vehicle (100) flies, as input,

- accessing the third marker data set (1423) stored in the memory unit (140) comprising at least one marker image for each said marker,

- generating the first prediction in which the marker in the image belongs to which of the marker images in the third marker dataset (1423) is predicted by applying the first classification algorithm to the image,

- generating the second prediction in which the marker in the image belongs to which of the marker images in the third marker dataset (1423) is predicted by applying the second classification algorithm, the type of which is different from that of the said first classification algorithm, to the image,

- generating the third prediction in which the marker in the image belongs to which of the images in the third marker dataset (1423) is predicted by applying the third classification algorithm, the type of which is different from those of the said first classification algorithm and the second classification algorithm, to the image,

- ensuring that the marker in the image is matched to the marker image with the highest number of predictions as a result of the said first prediction, the said second prediction, and the said third prediction,

- accessing the setting map (143) and being configured to determine the position of the matched marker on the third axis (403) on the setting map.