Granted Patent B2

US 9,852,631 · App. 15/060,830 · Granted Dec 26, 2017

Abstract

Claims (64)

1. A method for parking occupancy detection comprising: collecting video of a blockface with at least one video recording module; identifying a number of possible parking spaces along said blockface in said collected video; defining a region of interest for each of said possible parking spaces; detecting a time dependent occupancy of said defined regions of interest for each of said possible parking spaces; and reporting said time dependent occupancy; wherein defining a region of interest for each of said possible parking spaces further comprises: introducing a line path in a street, parallel to a traffic flow and a selected distance from a curb, wherein said fine path avoids parked vehicles and wherein said possible parking spaces are all occupied in said video frame; introducing a Sine path outside of said street, parallel to said traffic flow and a selected distance from said curd, wherein said line path avoids parked vehicles and wherein said possible parking spaces are all occupied in said video frame; introducing a Sine path along said curb in a video frame wherein said possible parking spaces are substantially unoccupied in said video frame: and introducing a dividing line between each of said possible parking spaces.

2. The method of claim 1 further comprising:

identifying an optimal location for said at least one video recording module;

configuring said at least one video recording module in said optimal location; and

adjusting said at least one video recording module to optimally observe said blockface.

3. The method of claim 1 wherein identifying a number of possible parking spaces along said blockface in said collected video further comprises:

indexing said possible parking spaces along said blockface.

4. The method of claim 1 further comprising:

performing drift correction on said collected video of said blockface.

5. The method of claim 1 wherein detecting occupancy of said defined regions of interest for each of said possible parking spaces further comprises:

performing an image-based parked vehicle detection comprising at east one of:

a spatiotemporal vehicle detection; and

a computer vision based vehicle detection.

6. The method of claim 5 further comprising:

fusing said image-based parked vehicle detection comprising said spatiotemporal vehicle detection and said image-based parked vehicle detection comprising said computer vision based vehicle detection.

7. The method of claim 1 wherein collecting video of a blockface with at least one video recording module comprises collecting video with a plurality of video recording modules, said method further comprising:

merging said time dependent detected occupancy of said defined regions of interest for each of said possible parking spaces provided from each of said plurality of video recording modules.

8. The method of claim 1 further comprising:

performing a ground truth detection in order to collect a true occupancy of said possible parking spaces.

9. A system for parking occupancy detection comprising: a video recording module configured to collect image data of a blockface; a processor; and a computer-usable medium embodying computer code, said computer-usable medium being coupled to said processor, said computer code comprising non-transitory instruction media executable by said processor configured for: identifying a number of possible parking spaces along said blockface in said collected video; defining a region of interest for each of said possible parking spaces; detecting a time dependent occupancy of said defined regions of interest for each of said possible parking spaces; and reporting said time dependent occupancy; wherein defining a region of interest for each of said possible parking spaces further comprises: introducing a line path in a street, parallel to a traffic flow and a selected distance from a curb, wherein said line path avoids parked vehicles and wherein said possible parking spaces are all occupied in said video frame; introducing a line path outside of said street, parallel to said traffic flow and a selected distance from said curd, wherein said line path avoids parked vehicles and wherein said possible parking spaces are all occupied in said video frame; introducing a Sine path along said curb in a video frame wherein said possible parking spaces are substantially unoccupied in said video frame: and introducing a dividing line between each of said possible parking spaces.

10. The system of claim 9 wherein said video recording module further comprises:

a video recorder affixed to a video trailer said video trailer comprising:

a vertically extending mast configured to hold said video recorder; and

a power source for powering said video recorder.

11. The system of claim 9 wherein identifying a number of possible parking spaces along said blockface in said collected video further comprises:

indexing said possible parking spaces along said blockface.

12. The system of claim 9 wherein said computer code comprising non-transitory instruction media executable by said processor is further configured for:

performing drift correction on said collected video of said blockface.

13. The system of claim 9 wherein detecting occupancy of said defined regions of interest for each of said possible parking spaces further comprises:

performing an image-based parked vehicle detection comprising at east one of:

a spatiotemporal vehicle detection; and

a computer vision based vehicle detection; and

fusing said image-based parked vehicle detection comprising said spatiotemporal vehicle detection and said image-based parked vehicle detection comprising said computer vision based vehicle detection.

14. The system of claim 9 wherein collecting video of a blockface with at least one video recording module comprises collecting video with a plurality of video recording modules, said method further comprising:

merging said time dependent detected occupancy of said defined regions of interest for each of said possible parking spaces provided from each of said plurality of video recording modules.

15. The system of claim 9 wherein said computer code comprising non-transitory instruction media executable by said processor is further configured for:

providing a ground truth module comprising a graphical user interface for performing a ground truth detection in order to collect a true occupancy of said possible parking spaces.

16. A method for parking occupancy detection comprising:

identifying an optimal location for at least one video recording module;

configuring said at least one video recording module in said optimal location;

adjusting said at least one video recording module to optimally observe said blockface;

indexing said possible parking spaces along said blockface;

collecting video of a blockface with at least one video recording module;

identifying a number of possible parking spaces along said blockface in said collected video;

defining a region of interest for each of said possible parking spaces wherein defining a region of interest further comprises:

introducing a line path in a street, parallel to a traffic flow and a selected distance from a curb, wherein said line path avoids parked vehicles and wherein said possible parking spaces are all occupied in said video frame;

introducing a line path outside of said street, parallel to said traffic flow and a selected distance from said curb, wherein said line path avoids parked vehicles and wherein said possible parking spaces are all occupied in said video frame;

introducing a line path along said curb in a video frame wherein said possible parking spaces are substantially unoccupied in said video frame; and

introducing a dividing line between each of said possible parking spaces;

detecting a time dependent occupancy of said defined regions of interest for each of said possible parking spaces;

performing a ground truth detection in order to collect a true occupancy of said possible parking spaces;

merging the time dependent occupancies from the multiple cameras observing each blockface; and

reporting said time dependent occupancy.

17. The method of claim 16 further comprising:

performing drift correction on said collected video of said blockface;

performing an image-based parked vehicle detection comprising at least one of:

a spatiotemporal vehicle detection; and

a computer vision based vehicle detection; and

fusing said image-based parked vehicle detection comprising said spatiotemporal vehicle detection and said image-based parked vehicle detection comprising said computer vision based vehicle detection.

18. The method of claim 16 wherein defining a region of interest for each of said possible parking spaces further comprises:

introducing a line path in a street, parallel to a traffic flow and a selected distance from a curb, wherein said line path avoids parked vehicles and wherein said possible parking spaces are all occupied in said video frame;

introducing a line path outside of said street, parallel to said traffic flow and a selected distance from said curb, wherein said line path avoids parked vehicles and wherein said possible parking spaces are all occupied in said video frame;

introducing a line path along, said curb in a video frame wherein said possible parking spaces are substantially unoccupied in said video frame; and

introducing a dividing line between each of said possible parking spaces.

Assignments (8)