Image processing method and apparatus implementing the same
An image processing method and a device configured to implement the same are disclosed. The method comprises: from an imaging device, obtaining image data that comprises temporally consecutive image frames; performing feature extraction on each of the obtained image frames; dynamically retaining extracted feature data of the obtained image frames in a feature accumulation database by regulating data retention in the feature accumulation database to a selective subset of the extracted feature data from the obtained image frames; performing Random Sample Consensus (RANSAC) operation on the selective subset of the extracted feature data from the feature accumulation database; and generating an estimation model from output of the RANSAC operation based on at least one of an extracted feature data of a current image frame or extracted feature data of one or more temporally preceding image frames of the obtained image frames.
1. A computer implemented image processing method, comprising:
from an imaging device, obtaining image data that comprises temporally consecutive image frames;
performing feature extraction on each of the obtained image frames;
dynamically retaining extracted feature data of the obtained image frames in a feature accumulation database by regulating data retention in the feature accumulation database to a selective subset of the extracted feature data from the obtained image frames;
performing Random Sample Consensus (RANSAC) operation on the selective subset of the extracted feature data from the feature accumulation database; and
generating an estimation model from output of the RANSAC operation based on at least one of an extracted feature data of a current image frame or extracted feature data of one or more temporally preceding image frames of the obtained image frames.
2. The method of claim 1 , wherein performing RANSAC operation on the selective subset of the extracted feature data comprises:
among the obtained image frames, determining whether a newly obtained image frame constitutes a key frame, and
forwarding the extracted feature data of the key frame from the feature accumulation database for selective RANSAC operation.
3. The method of claim 2 , further comprising:
further forwarding an accumulated extracted feature data preceding that of the key frame from the feature accumulation database for selective RANSAC operation.
4. The method of claim 2 , wherein the key frame determination process comprises:
performing consistency check on the extracted feature data from the current image frame with respect to a temporally preceding estimation model, and
when a consistency rating of the extracted feature data of the current image frame is below a consistency threshold, designating the current image frame as a key frame.
5. The method of claim 4 , wherein when the current image frame is designated as the key frame, retaining the extracted feature data of the key frame in the feature accumulation database for a first temporal duration;
wherein when the current image frame is not designated as the key frame, retaining the extracted feature data of the current image frame in the feature accumulation database for a second temporal duration that is shorter than the first temporal duration.
6. The method of claim 4 , further comprising:
for each of the generated estimation model, evaluating an estimation model quality by generating a model quality measure variable,
wherein the model quality measure variable comprises a function of RANSAC support votes, total number of features used for RANSAC, and current model residual error.
7. The method of claim 6 , wherein dynamically retaining extracted feature data of the obtained image frames in a feature accumulation database comprises:
regulating a discard rate of the retained extracted feature data of the obtained image frames in the feature accumulation database based at least in part on the model quality measure variable.
8. The method of claim 7 , wherein regulating a discard rate of the feature accumulation database comprises:
generating a control variable having a value not greater than 1 by inputting the model quality measure variable into a dynamic accumulation control function,
wherein the dynamic accumulation control function comprises a monotonically decreasing function.
9. The method of claim 8 , further comprising:
when the dynamic accumulation control function satisfies a predetermined first quality threshold, setting the control variable to a value less than 1; and
reducing a retention volume of extracted feature data in the feature accumulation database by:
reducing a current data accumulation weight factor by factor of the control variable, and
removing retained extracted feature data in the feature accumulation database with weight factor less than a predetermined second quality threshold.
10. The method of claim 9 , wherein the dynamic accumulation control function is designated to affect an inversely proportional relationship between the estimation model quality and the control variable.
11. A computer implemented image processing method, comprising:
from an imaging device, obtaining image data that comprises temporally consecutive image frames;
performing feature extraction on each of the obtained image frames;
accumulating extracted feature data of the obtained image frames in a feature accumulation database; and
generating an estimation model by performing RANSAC operation on extracted feature data of a selective subset of the temporally consecutive image frames,
wherein performing RANSAC operation on extracted feature data of a selective subset comprises:
among the obtained image frames, determining whether a newly obtained image constitutes a key frame, and
forwarding the extracted feature data of the key frame from the feature accumulation database for selective RANSAC operation,
wherein the key frame determination process comprises:
performing consistency check on the extracted feature data from a current image frame with respect to a temporally preceding estimation model, and
when a consistency rating of the extracted feature data of the current image frame is below a consistency threshold, designating the current image frame as a key frame,
wherein when the current image frame is designated key frame, retaining the extracted feature data of the key frame in the feature accumulation database for a first temporal duration;
wherein when the current image frame is not designated key frame, retaining the extracted feature data of the current image frame in the feature accumulation database for a second temporal duration that is shorter than the first temporal duration.
12. A computer implemented image processing method, comprising:
from an imaging device, obtaining image data that comprises temporally consecutive image frames;
performing feature extraction on each of the obtained image frames;
accumulating extracted feature data of the obtained image frames in a feature accumulation database;
generating an estimation model by performing RANSAC operation on extracted feature data of a selective subset of the temporally consecutive image frames; and
for each of the generated estimation model, evaluating an estimation model quality by generating a model quality measure variable,
wherein the model quality measure variable comprises a function of RANSAC support votes, total number of features used for RANSAC, and current model residual error,
wherein performing RANSAC operation on extracted feature data of a selective subset comprises:
among the obtained image frames, determining whether a newly obtained image constitutes a key frame, and
forwarding the extracted feature data of the key frame from the feature accumulation database for selective RANSAC operation,
wherein the key frame determination process comprises:
performing consistency check on the extracted feature data from a current image frame with respect to a temporally preceding estimation model, and
when a consistency rating of the extracted feature data of the current image frame is below a consistency threshold, designating the current image frame as a key frame.
13. The method of claim 12 , wherein accumulating extracted feature data of the obtained image frames in a feature accumulation database comprises:
regulating a discard rate of the accumulated extracted feature data in the feature accumulation database based at least in part on the model quality measure variable.
14. The method of claim 13 , wherein regulating a discard rate of the feature accumulation database comprises:
generating a control variable having a value not greater than 1 by inputting the model quality measure variable into a dynamic accumulation control function,
wherein the dynamic accumulation control function comprises a monotonically decreasing function.
15. The method of claim 14 , further comprising:
when the dynamic accumulation control function does not satisfy a predetermined first quality threshold, setting the control variable to a value of 1; and
expanding a retention volume of extracted feature data in the feature accumulation database by accumulating extracted feature data from the current image frame into the feature accumulation database.
16. The method of claim 15 , wherein the dynamic accumulation control function is designated to affect an inversely proportional relationship between the estimation model quality and the control variable.
17. An image processing system, comprising:
a processing device configured to obtain image data that comprises temporally consecutive image frames from an imaging device, the processing device comprising:
a feature extraction component configured to receive image data from the imaging device and to extract feature data from each of the obtained image frames;
a feature accumulation database in data communication with the feature extraction component, configured to selectively retain the extracted feature data;
a Random Sample Consensus (RANSAC) processing component configured to apply RANSAC process on the extracted feature data from the feature accumulation database;
a model generator arranged downstream of the RANSAC processing component, configured to generate estimated model from an output of the RANSAC processing component; and
a key frame selection component arranged communicatively between the feature accumulation database and the RANSAC processing component, configured to perform consistency check on the extracted feature data from a newly obtained image frame with respect to a temporally preceding estimation model from the model generator.
18. The system of claim 17 , wherein the key frame selection component is further configured to:
designate the newly obtained image frame as a key frame when a consistency rating of the extracted feature data of the newly obtained image frame is below a predetermined consistency threshold, and
forwarding an accumulated extracted feature data that includes the extracted feature data of the key frame from the feature accumulation database to the RANSAC processing component for selective RANSAC processing.
19. The system of claim 18 , further comprising a model quality measuring component arranged communicatively downstream of the model generator, configured to evaluate estimation model quality for each of the generated estimation model, and to generate a model quality measure variable.
20. The system of claim 19 , further comprising a dynamic feature accumulation controller configured to dynamically regulate data accumulation in the feature accumulation database in accordance with the model quality measure variable.