Method of mobile platform detecting and tracking dynamic objects and computer-readable medium thereof

1. A method of a mobile platform detecting and tracking moving objects, comprising:

acquiring a three-dimensional (3D) image of an environment in which the mobile platform moves;

extracting and removing, using a processor, a floor plane from the 3D image;

individually separating, using the processor, objects from the 3D image with the floor plane removed;

estimating movement of the respective separated objects; and

registering the respective separated objects as formal dynamic objects when a predetermined condition is satisfied,

wherein the predetermined condition entails a covariance value being less than a predetermined critical value.

2. The method according to claim 1 , wherein the extracting and removing the floor plane from the 3D image comprises using a random sample consensus (RANSAC) algorithm.

3. The method according to claim 1 , wherein the 3D image acquired by the mobile platform comprises a depth map acquired by a time-of-flight (TOF) sensor.

4. The method according to claim 3 , wherein the extracting and removing the floor plane from the 3D image comprises performing a RANSAC algorithm using pixel information of the depth map acquired by the TOF sensor as data.

5. The method according to claim 3 , wherein the individually separating the objects from the 3D image with the floor plane removed comprises using a blob algorithm.

6. The method according to claim 3 , wherein the movement of the respective separated objects is estimated using a joint probability data association filter (JPDAF).

7. The method according to claim 6 , further comprising registering the respective separated objects as formal dynamic objects when it is determined that the respective separated objects are dynamic objects as a result of estimating the movement of the respective separated objects using the JPDAF and when the predetermined condition is satisfied.

8. The method according to claim 7 , further comprising estimating location or velocity of the registered formal dynamic objects to track the dynamic objects.

9. The method according to claim 8 , further comprising designing a travel route along which the mobile platform avoids the registered formal dynamic objects using the estimated location or velocity of the registered formal dynamic objects.

10. At least one non-transitory computer readable medium comprising computer readable instructions that control at least one processor to implement a method, comprising:

acquiring a three-dimensional (3D) image of an environment in which the mobile platform moves;

extracting and removing a floor plane from the 3D image;

individually separating objects from the 3D image with the floor plane removed;

estimating movement of the respective separated objects; and

registering the respective separated objects as formal dynamic objects when a predetermined condition is satisfied,

wherein the predetermined condition entails a covariance value being less than a predetermined critical value.

11. The at least one non-transitory computer readable medium comprising computer readable instructions that control at least one processor to implement the method of claim 10 , wherein the extracting and removing the floor plane from the 3D image comprises using a random sample consensus (RANSAC) algorithm.

12. The at least one non-transitory computer readable medium comprising computer readable instructions that control at least one processor to implement the method of claim 10 , wherein the 3D image acquired by the mobile platform comprises a depth map acquired by a time-of-flight (TOF) sensor.

13. The at least one non-transitory computer readable medium comprising computer readable instructions that control at least one processor to implement the method of claim 12 , wherein the extracting and removing the floor plane from the 3D image comprises performing a RANSAC algorithm using pixel information of the depth map acquired by the TOF sensor as data.

14. The at least one non-transitory computer readable medium comprising computer readable instructions that control at least one processor to implement the method of claim 12 , wherein the individually separating the objects from the 3D image with the floor plane removed comprises using a blob algorithm.

15. The at least one non-transitory computer readable medium comprising computer readable instructions that control at least one processor to implement the method of claim 12 , wherein the movement of the respective separated objects is estimated using a joint probability data association filter (JPDAF).

16. The at least one non-transitory computer readable medium comprising computer readable instructions that control at least one processor to implement the method of claim 15 , further comprising registering the respective separated objects as formal dynamic objects when it is determined that the respective separated objects are dynamic objects as a result of estimating the movement of the respective separated objects using the JPDAF and when the predetermined condition is satisfied.

17. The at least one non-transitory computer readable medium comprising computer readable instructions that control at least one processor to implement the method of claim 16 , further comprising estimating location or velocity of the registered formal dynamic objects to track the dynamic objects.

18. The at least one non-transitory computer readable medium comprising computer readable instructions that control at least one processor to implement the method of claim 17 , further comprising designing a travel route along which the mobile platform avoids the registered formal dynamic objects using the estimated location or velocity of the registered formal dynamic objects.