Fingerprinting media content using hashing
A server device includes memory to store: a database of frame fingerprints ordered sequentially by media item number and frame number for media items, and a hash table containing (key, value) entries each including a hashed frame fingerprint and index number(s) to the database. A processor coupled to the memory is to, for an input set of hashed frame fingerprints, select a most recent hashed frame fingerprint. The processor performs a lookup with the most recent hashed frame fingerprint to identify a plurality of index numbers from a matching entry. The processor retrieves a target plurality of frame fingerprints from the database using each of the plurality of index numbers as a starting point; calculates a rank distance between the input set and each respective target plurality of frame fingerprints; and selects the target plurality of frame fingerprints for which the rank distance is shortest, to identify a media item.
1. A system comprising:
non-transitory memory storing instructions; and
data processing hardware in communication with the non-transitory memory and configured to execute the instructions, the executed instructions causing the data processing hardware to perform operations comprising:
receiving an input set of hashed frame fingerprints of an unidentified media item from a media device, the hashed frame fingerprints ordered chronologically by frame number in the input set, and the input set of hashed frame fingerprints comprising a most recent hashed frame fingerprint;
identifying an index number from a hash table that matches the most recent hashed fingerprint and a target frame fingerprint corresponding to the identified index number;
identifying a target block of frame fingerprints corresponding to the target frame fingerprint, the target block of frame fingerprints ordered chronologically sequentially backwards by frame number from the target frame fingerprint;
determining a rank distance corresponding to a distance between the input set of hashed frame fingerprints and the target block of frame fingerprints;
identifying a media item identifier corresponding to the target block of frame fingerprints associated with a shortest rank distance from the input set of hashed frame fingerprints, the identified media item identifier identifying the media item associated with the input set of hashed frame fingerprints; and
sending the identified media item identifier to the media device.
2. The system of claim 1 , wherein the media device is configured to request overlay content from an overlay decision engine for display on a viewer in communication with the media device based on the media item identifier.
3. The system of claim 1 , wherein the operations further comprise:
identifying a plurality of index numbers from the hash table that matches the most recent hashed fingerprint and the target frame fingerprint corresponding to each identified index number;
for each identified index number, determining the target block of frame fingerprints corresponding to the target frame fingerprint; and
determining the rank distance corresponding to the distance between the input set of hashed frame fingerprints and each determined target block of frame fingerprints.
4. The system of claim 1 , wherein a number of frame fingerprints within the target block of frame fingerprints corresponds to a number of the frame fingerprints in the input set of hashed frame fingerprints.
5. The system of claim 1 , wherein determining the rank distance from the input set of hashed frame fingerprints and the target block of frame fingerprints comprises calculating a sum of an absolute difference between ordinal values of respective hashed frame fingerprints of the input set and the frame fingerprints for the target block.
6. The system of claim 1 , wherein the operations further comprise:
when identifying the media item identifier, retrieving a frame number for the target frame fingerprint from the target block of frame fingerprints associated with the shortest rank distance; and
when sending the identified media item identifier, sending the frame number for the target frame fingerprint to the media device.
7. The system of claim 1 , wherein the operations further comprise storing frame fingerprints ordered sequentially by media and frame number for a plurality of media items in a database.
8. The system of claim 7 , wherein the plurality of media items comprise videos or television programs.
9. The system of claim 1 , wherein the operations further comprise storing a hash table containing a plurality of (key, value) pair entries, each entry comprising a hashed frame fingerprint and one or more index numbers, each index number mapping the hashed frame fingerprint to a corresponding slot in a database.
10. The system of claim 9 , wherein the operation of identifying the index number from the hash table further comprises identifying the index number from the plurality of (key, value) pair entries.
11. A method comprising:
receiving, at data processing hardware, an input set of hashed frame fingerprints of an unidentified media item from a media device, the hashed frame fingerprints ordered chronologically by frame number in the input set, and the input set of hashed frame fingerprints comprising a most recent hashed frame fingerprint
identifying, by the data processing hardware, an index number from a hash table that matches the most recent hashed fingerprint and a target frame fingerprint corresponding to the identified index number;
identifying, by the data processing hardware, a target block of frame fingerprints corresponding to the target frame fingerprint, the frame fingerprints ordered chronologically sequentially backwards by frame number from the target frame fingerprint;
determining, by the data processing hardware, a rank distance corresponding to a distance between the input set of hashed frame fingerprints and the target block of frame fingerprints;
identifying, by the data processing hardware, a media item identifier corresponding to the target block of frame fingerprints associated with a shortest rank distance from the input set of hashed frame fingerprints, the identified media item identifier identifying the media item associated with the input set of hashed frame fingerprints; and
sending, by the data processing hardware, the identified media item identifier to the media device.
12. The method of claim 11 , wherein the media device is configured to request overlay content from an overlay decision engine for display on a viewer in communication with the media device based on the media item identifier.
13. The method of claim 11 further comprising:
identifying, by the data processing hardware, a plurality of index numbers from the hash table that matches the most recent hashed fingerprint and the target frame fingerprint corresponding to each identified index number;
for each identified index number, determining, by the processing, the target block of frame fingerprints corresponding to the target frame fingerprint; and
determining, by the data processing hardware, the rank distance corresponding to the distance between the input set of hashed frame fingerprints and each determined target block of frame fingerprints.
14. The method of claim 11 , wherein a number of frame fingerprints within the target block of frame fingerprints corresponds to a number of the frame fingerprints in the input set of hashed frame fingerprints.
15. The method of claim 11 , wherein determining the rank distance from the input set of hashed frame fingerprints and the target block of frame fingerprints comprises calculating a sum of an absolute difference between ordinal values of respective hashed frame fingerprints of the input set and the frame fingerprints for the target block.
16. The method of claim 11 , further comprising:
when identifying the media item identifier, retrieving, by the data processing hardware, a frame number for the target frame fingerprint from the target block of frame fingerprints associated with the shortest rank distance; and
when sending the identified media item identifier, sending, from the data processing hardware, the frame number for the target frame fingerprint to the media device.
17. The method of claim 11 , wherein the frame fingerprints correspond to media items comprising a video or television program.
18. The method of claim 11 , wherein receiving the input set of hashed frame fingerprints further comprises receiving the input set of hashed frame fingerprints from an automatic content recognition (ACR) device, the ACR device configured to:
receive a media feed, the media feed comprising the unidentified media item from the media device;
fingerprint the unidentified media item; and
apply a hash function to the fingerprinted unidentified media item.
19. The method of claim 11 , wherein receiving the input set of hashed frame fingerprints further comprises receiving the input set of hashed frame fingerprints from an automatic content recognition (ACR) device, the ACR device configured to generate hashed frame fingerprints by applying a hash function to frame fingerprints of the unidentified media item from the media device.
20. The method of claim 11 , wherein receiving the input set of hashed frame fingerprints further comprises receiving the input set of hashed frame fingerprints from an automatic content recognition (ACR) device, the ACR device configured to fingerprint a frame of the unidentified media item by discrete cosine transform.