Available parking space detection localization using historical aggregation shifting

1. A method for parking spot localization, the method comprising:

receiving, by processing circuitry and from at least one ascertaining vehicle driving through a street segment, data from at least one sensor of the at least one ascertaining vehicle corresponding to detected objects in the street segment;

determining, by the processing circuitry, a detected object distribution based on the received data;

shifting, by the processing circuitry, the detected object distribution based on a predefined parking region distribution and the detected object distribution; and

generating, by the processing circuitry, a parking map of available parking spots on the street segment based on the shifted detected object distribution;

wherein:

(I) the parking region distribution is a histogram having:

(1) a bin width (W);

(2) a length (L) corresponding to a length of the street segment;

(3) a height of one (1) inside allowed parking regions; and

(4) a height of zero (0) outside allowed parking regions; and/or

(II) the method further comprises the processing circuitry calculating, based on a score function, a score corresponding to the parking region distribution and the detected object distribution, the shift of the detected object distribution is based on a minimization of the calculated score, the method further comprises the processing circuitry calculating a regularized score corresponding to the parking region distribution and the detected object distribution, and the regularized score is calculated using a Ridge Regression model.

2. The method of claim 1 , wherein the receiving of the data, the determining of the detected object distribution, and the shifting of the detected object distribution are performed each time at least one of the at least one ascertaining vehicle drives through the street section.

3. The method of claim 1 , wherein the data are ascertained using distance sensors situated on the at least one ascertaining vehicle.

4. The method of claim 1 , wherein the detected object distribution is a second histogram having 1) the bin width (W); 2) the length (L) corresponding to the length of the street segment; 3) a height of one (1) where an object is detected; and 4) a height of zero (1) where no object is detected.

5. The method of claim 4 , wherein:

the method comprises the calculation of feature (II) of calculating, based on the score function, the score corresponding to the parking region distribution and the detected object distribution;

the score function is defined as

Σ

⁢

(

T

i

-

O

i

)

2

L

;

T i represents an i-th bin from the parking region histogram;

O i represents an i-th bin from the detected object histogram;

L represents the length of the street segment; and

the score function is calculated over the length of the street segment.

6. The method of claim 4 , wherein the bin width is equal to 1 meter.

7. A method for parking spot localization, the method comprising:

receiving, by processing circuitry and from at least one ascertaining vehicle driving through a street segment, data from at least one sensor of the at least one ascertaining vehicle and corresponding to detected objects in the parking region;

determining, by the processing circuitry, a detected object distribution based on the received data;

shifting, by the processing circuitry, the detected object distribution based on a historical distribution of objects detected in the street segment and a predefined parking region distribution; and

generating, by the processing circuitry, a parking map of available parking spots on the street segment based on the shifted detected object distribution;

wherein:

(I) the parking region distribution is a histogram having:

(1) a bin width (W);

(2) a length (L) corresponding to a length of the street segment;

(3) a height of one (1) inside allowed parking regions; and

(4) a height of zero (0) outside allowed parking regions; and/or

(II) the method further comprises the processing circuitry calculating, based on a score function, a score corresponding to the parking region distribution and the detected object distribution, the shift of the detected object distribution is based on a minimization of the calculated score, the method further comprises the processing circuitry calculating a regularized score corresponding to the parking region distribution and the detected object distribution, and the regularized score is calculated using a Ridge Regression model.

8. The method of claim 7 , wherein the shifting of the detected object distribution is performed each time at least one of the at least one ascertaining vehicle drives through the street section.

9. A system for parking spot localization, the system comprising:

processing circuitry, wherein the processing circuitry is configured to perform the following:

receive, from at least one ascertaining vehicle driving through a street segment, data from at least one sensor of the at least one ascertaining vehicle corresponding to detected objects in the street segment;

determine a detected object distribution based on the received data;

shift, by the processing circuitry, the detected object distribution based on a parking region distribution and the detected object distribution; and

generate, by the processing circuitry, a parking map of available parking spots on the street segment based on the parking region distribution and the shifted detected object distribution;

wherein:

(I) the parking region distribution is a histogram having:

(1) a bin width (W);

(2) a length (L) corresponding to a length of the street segment;

(3) a height of one (1) inside allowed parking regions; and

(4) a height of zero (0) outside allowed parking regions; and/or

(II) the processing circuitry is configured to calculate, based on a score function, a score corresponding to the parking region distribution and the detected object distribution, the shift of the detected object distribution is based on a minimization of the calculated score, the processing circuitry is configured to calculate a regularized score corresponding to the parking region distribution and the detected object distribution, and the regularized score is calculated using a Ridge Regression model.

10. The server system of claim 9 , wherein the receipt of the data, the determining of the detected object distribution, and the shifting of the detected object distribution are performed each time at least one of the at least one ascertaining vehicle drives through the street section.

11. A non-transitory computer-readable medium on which are stored instructions that are executable by a computer processor and that, when executed by the processor, cause the processor to perform to perform a method for parking spot localization, the method comprising:

receiving, from at least one ascertaining vehicle driving through a street segment, data from at least one sensor of the at least one ascertaining vehicle corresponding to detected objects in the street segment;

determining, by the processing circuitry, a detected object distribution based on the received data;

shifting, by the processing circuitry, the detected object distribution based on a parking region distribution and the detected object distribution; and

generating, by the processing circuitry, a parking map of available parking spots on the street segment based on the parking region distribution and the shifted detected object distribution;

wherein:

(I) the parking region distribution is a histogram having:

(1) a bin width (W);

(2) a length (L) corresponding to a length of the street segment;

(3) a height of one (1) inside allowed parking regions; and

(4) a height of zero (0) outside allowed parking regions; and/or

(II) the method further comprises the processing circuitry calculating, based on a score function, a score corresponding to the parking region distribution and the detected object distribution, the shift of the detected object distribution is based on a minimization of the calculated score, the method further comprises the processing circuitry calculating a regularized score corresponding to the parking region distribution and the detected object distribution, and the regularized score is calculated using a Ridge Regression model.