Image processing system and method
There is provided an image processing system and method for identifying a user. The system comprises a processor configured to identify a first user in an image, determine a plurality of characteristic vectors associated with the first user, compare the characteristic vectors associated with the first user with a plurality of predetermined characteristic vectors associated with a plurality of users including the first user, and identify the first user based on the comparison.
1 . An image processing system for identifying an entity, the system comprising means for:
a. determining a region within a received image of an entity wherein the region encloses the entity;
b. segmenting the region into one or more sub regions;
c. determining a first set of embedding vectors comprising a respective embedding vector for each of the sub regions, wherein each of the first set of embedding vectors is defined by a plurality of characteristic feature values that uniquely describe each sub region;
d. determining a second set of embedding vectors, wherein each of the second set of embedding vectors is associated with a respective one of a plurality of known entities, and each of the plurality of known entities is associated with a unique identifier;
e. comparing the first set of embedding vectors with the second set of embedding vectors; and
f. based on the comparison, determining whether the entity is matched with a matching known entity of the plurality of known entities or is a new entity, wherein:
i. if the entity is matched with the matching known entity, the entity is associated with the unique identifier associated with the matching known entity, or;
ii. if the entity is the new entity, the entity is associated with a new identifier.
2 . The system of claim 1 , further comprising means for authorising the matching known entity for entry or exit via a gate, and further comprising means for associating the unique identifier with a passenger related information or a bag tag number.
3 . The system of claim 1 , wherein the first set of embedding vectors are determined based on a first image of the entity and a further set of embedding vectors are determined based on a second image of the entity.
4 . The system of claim 3 , further comprising means for selecting a subset of optimum embedding vectors for the entity from the first set of embedding vectors and further sets of embedding vectors by identifying embedding vectors that have a largest value of a predetermined characteristic feature value.
5 . The system of claim 1 , wherein the plurality of characteristic feature values are associated with one or more of: biometric data, face features, height, style, clothing, pose, gender, age, emotion, destination gate, or gesture recognition.
6 . The system of claim 3 , wherein the system further comprises means for associating the first image with a first predetermined location and associating the second image with a second predetermined location different from the first predetermined location, wherein the first predetermined location and the second predetermined location are each associated with one or more of customer car parks, airport terminal entrances and exits, airline check-in areas, check-in kiosks, terminal concourses, customer shopping and/or dining areas, passenger lounges, security and passport control areas, customs and excise areas, arrival lounges, departure lounges, and baggage processing areas.
7 . The system of claim 1 further comprising means for associating latitude, longitude and timestamp data with a location of the entity in the received image.
8 . The system of claim 1 , wherein the entity is a user and the one or more sub regions includes a first sub region associated with a head of the user, a second sub region associated with a body of the user, and a third sub region associated with belongings accompanying the user, and wherein characteristic feature values are associated with one or more of:
biometric data, face features, height, style, clothing, pose, gender, age, emotion, destination gate, or gesture recognition.
9 . An image processing method for identifying an entity, the method comprising the steps of:
a. receiving an image of an entity and determining a region within the image that encloses the entity;
b. segmenting the region into one or more sub regions;
c. determining a first set of embedding vectors comprising a respective embedding vector for each of the one or more sub regions, wherein each of the first set of embedding vectors is defined by a plurality of characteristic feature values that uniquely describe each sub region;
d. determining a second set of embedding vectors, wherein each of the second set of embedding vectors is associated with a respective one of a plurality of known entities, and each of the plurality of known entities is associated with a unique identifier;
e. comparing the first set of embedding vectors with the second set of embedding vectors;
f. determining, based on the comparison, whether the entity is matched with a matching known entity of the plurality of known entities or is a new entity, wherein
i. if the entity is matched with the matching known entity, the entity is associated with the unique identifier associated with the matching known entity, or;
ii. if the entity is the new entity, the entity is associated with a new identifier.
10 . The method of claim 9 , further comprising authorising the matching known entity for entry or exit via a gate, and further comprising sending a message to actuate one or more infrastructure systems if any of the characteristic feature values exceeds a threshold value.
11 . The method of claim 10 , wherein the one or more infrastructure systems comprise one or more of: security barriers, public address systems, or emergency lighting systems.
12 . The method of claim 9 , further comprising associating the unique identifier with passenger related information or a bag tag number.
13 . The method of claim 9 , wherein determining whether the entity is matched with the matching known entity is based on a degree of similarity between each of the first set of embedding vectors and each of the second set of embedding vectors.
14 . The method of claim 9 , further comprising pre-processing the received image, wherein pre-processing comprises one or more of: sampling raw data, reducing background noise in the received image, defining a region of interest within each image, removing a background of the received image, and synchronising cameras.
15 . The method of claim 13 , further comprising determining a confidence score based on the degree of similarity, and/or flight related information associated with the unique identifier associated with the matching known entity.
16 . The method of claim 9 , wherein the entity is a user and the one or more sub regions includes a first sub region associated with a head of the user, a second sub region associated with a body of the user, and a third sub region associated with belongings accompanying the user, and wherein characteristic feature values are associated with one or more of: biometric data, face features, height, style, clothing, pose, gender, age, emotion, destination gate, or gesture recognition.
17 . The method of claim 9 , further comprising associating latitude, longitude and timestamp data with a location of the entity in the received image.