Feature trackability ranking, systems and methods

A computer-based method for establishing a trackability ranking order from tracked observed image features using at least one processor, the method comprising:

• establishing a tracking region within a first image sequence using a feature detection algorithm;
• compiling groups of tracked image features within the tracking region using a feature tracking algorithm;
• assigning saliency metrics to at least two of the groups of the tracked image features based on one or more feature property measurements within the tracking region;
• determining a trackability ranking order for the tracked image features within the first image sequence using the trackability ranking algorithm, wherein the trackability ranking algorithm comprises at least sorting the groups of the tracked image features by a result of a linear combination of the saliency metrics;
• and
• selecting a subset of the tracked image features according to the trackability ranking order to facilitate an image recognition-based operation within a second image sequence according to the trackability ranking order using the trackability ranking algorithm.

A computer program product embedded in a non-transitory computer readable medium comprising instructions executable by a computer processor for establishing a trackability ranking order from tracked observed image features, the instructions being executable by a computer processor to execute processing comprising:

• establishing a tracking region within a first image sequence using a feature detection algorithm;
• compiling groups of tracked image features within the tracking region using a feature tracking algorithm;
• assigning saliency metrics to at least two of the groups of the tracked image features based on one or more feature property measurements within the tracking region;
• determining a trackability ranking order for the tracked image features within the first image sequence using the trackability ranking algorithm, wherein the trackability ranking algorithm comprises at least sorting the groups of the tracked image features by a result of a linear combination of the saliency metrics;
• and
• selecting a subset of the tracked image features according to the trackability ranking order to facilitate an image recognition-based operation within a second image sequence according to the trackability ranking order using the trackability ranking algorithm.

A vehicle, comprising:

• one or more optical sensors configured to capture a first image sequence;
• and
• at least one tracking analysis device comprising a processor communicatively coupled with the one or more optical sensors and a database storing trackability ranking software code that, when executed by the processor, causes the processor to:
• establish a tracking region within the first image sequence using a feature detection algorithm;
• compile groups of tracked image features within the tracking region using a feature tracking algorithm;
• assign saliency metrics to at least two of the groups of the tracked image features based on one or more feature property measurements within the tracking region;
• determine a trackability ranking order for the tracked image features within the first image sequence using the trackability ranking algorithm, wherein the trackability ranking algorithm comprises at least sorting the groups of the tracked image features by a result of a linear combination of the saliency metrics;
• and
• select a subset of the tracked image features according to the trackability ranking order to facilitate an image recognition-based operation within a second image sequence according to the trackability ranking order using the trackability ranking algorithm.

A server comprising:

• a processor communicatively coupled with a database storing trackability ranking software code that, when executed by the processor, causes the processor to:
• install the trackability ranking software code on at least one tracking device such that, when executed by the at least one tracking device, the trackability ranking software code causes the at least one tracking device to:
• establish a tracking region within a first image sequence using a feature detection algorithm;
• compile groups of tracked image features within the tracking region using a feature tracking algorithm;
• assign saliency metrics to at least two of the groups of the tracked image features based on one or more feature property measurements within the tracking region;
• determine a trackability ranking order for the tracked image features within the first image sequence using the trackability ranking algorithm, wherein the trackability ranking algorithm comprises at least sorting the groups of the tracked image features by a result of a linear combination of the saliency metrics;
• and
• select a subset of the tracked image features according to the trackability ranking order to facilitate an image recognition-based operation within a second image sequence according to the trackability ranking order using the trackability ranking algorithm.