Smart door system
View Patent ↗A method for smart door operation implemented on a computer may include obtaining image data of one or more subjects. The image data may be acquired by one or more cameras in communication with the computer. The method may also include determining, based on the image data, one or more gait features of at least one of the one or more subjects and determining, at least in part based on the one or more gait features, identity information of the at least one of the one or more subjects. The method may further include sending an unlocking signal to a lock based on the unlocking signal so that the lock is unlocked.
1. A method implemented on a computer for smart door operation via the computer, the method comprising:
obtaining image data of one or more subjects, the image data being acquired by one or more cameras in communication with the computer;
determining, based on the image data, one or more gait features of at least one of the one or more subjects;
determining, based on the gait features of at least one of the one or more subjects, a gait recognition result indicating whether the at least one of the one or more subjects is consistent with one of one or more allowed entry identities;
determining, based on facial features of at least one of the one or more subjects derived from the image data, a facial recognition result indicating whether the at least one of the one or more subjects is consistent with the one of one or more allowed entry identities;
in response to determining that the facial recognition result is different from the gait recognition result, updating gait features or facial features of the one of one or more allowed entry identities based on the facial recognition result and the gait recognition result; and
sending, based on identity information of the at least one of the one or more subjects, an unlocking signal to a lock, so that the lock is unlocked.
2. The method of claim 1 , wherein the determining the gait recognition result includes:
determining, based on the one or more gait features using a gait recognition model, a first confidence level that the identity information of the at least one of the one or more subjects corresponds to one of one or more allowed entry identities;
determining the gait recognition by determining whether the first confidence level exceeds or equals a first threshold.
3. The method of claim 2 , wherein the gait recognition model is determined via a process including:
obtaining video data associated with the one or more allowed entry identities as the training samples;
extracting gait features from each of the training samples; and
training the gait recognition model based on the gait features of each of the training samples.
4. The method of claim 1 , wherein the determining the facial recognition result includes:
determining, based on the one or more facial features, a second confidence level that the identity information of the at least one of the one or more subjects corresponds to the one of the one or more allowed entry identities;
determining the facial recognition result by determining whether the second confidence level exceeds or equals a second threshold.
5. The method of claim 4 , wherein the determining, based on the one or more facial features, a second confidence level that the identity information is the one of the one or more allowed entry identities includes:
obtaining a face recognition model; and
determining the second confidence level based on the facial features of the at least one of the one or more subjects using the face recognition model.
6. The method of claim 5 , wherein the face recognition model is determined via a process including:
obtain facial images of the one or more allowed entry identities as training samples;
extracting facial features from each of the training samples; and
training the face recognition model based on the facial features of each of the training samples.
7. The method of claim 1 , wherein the gait features include at least one of:
a swinging amplitude of a leg, a swinging frequency of the leg, a swinging phase of the leg, a swinging amplitude of a waist, a swinging frequency of the waist, a swinging phase of the waist, a linear velocity of a joint, an angular velocity of the joint, a step size, a stride frequency, a target velocity, or a target acceleration.
8. The method of claim 1 , wherein the facial features include at least one of:
an eye center point, a nose tip, a lip corner point, a facial contour point, an eye contour point, a nose contour point, a lip contour point, or an eyebrow contour point.
9. The method of claim 1 , wherein the method further includes:
showing a message to the at least one of the one or more subjects based on the identity information of the at least one of the one or more subjects.
10. The method of claim 9 , wherein the message includes a preset question, and the method further includes:
verifying the identity information of the at least one of the one or more subjects based on a response of the at least one of the one or more subjects to the preset question.
11. The method of claim 1 , wherein the method further includes:
obtaining information associated with the one of the at least one of the one or more subjects;
determining event information associated with the at least one of the one or more subjects based on current entry and exit information of the at least one of the one or more subjects and the information associated with the at least one of the one or more subjects; and
showing the event information to the at least one of the one or more subjects.
12. The method of claim 1 , wherein the method further includes:
obtaining setting information of a device in a building associated with the at least one of the one or more subjects; and
operating the device based on the setting information of the device.
13. The method of claim 1 , wherein the method further includes:
obtaining personalized information associated with the one of the one or more allowed entry identities; and
showing the personalized information to the at least one of the one or more subjects.
14. The method of claim 1 , wherein the method further includes:
sending information associated with the at least one of the one or more subjects to a client terminal based on the identity information;
generating a warning signal for the at least one of the one or more subjects based on the identity information;
sending the information associated with the at least one of the one or more subjects to a client terminal associated with a specific contact based on the identity information;
reporting the police based on the identity information; or
recording the information associated with the at least one of the one or more subjects.
15. A smart door apparatus, comprising at least one processor and at least one storage;
the at least one storage being configured to store an instruction; and
the processor being configured to execute the instructions to implement a method for smart door operation, the method comprising:
obtaining image data of one or more subjects, the image data being acquired by one or more cameras in communication with the computer;
determining, based on the image data, one or more gait features of at least one of the one or more subjects;
determining, based on the gait features of at least one of the one or more subjects, a gait recognition result indicating whether the at least one of the one or more subjects is consistent with one of one or more allowed entry identities;
determining, based on facial features of at least one of the one or more subjects derived from the image data, a facial recognition result indicating whether the at least one of the one or more subjects is consistent with the one of one or more allowed entry identities;
in response to determining that the facial recognition result is different from the gait recognition result, updating gait features or facial features of the one of one or more allowed entry identities based on the facial recognition result and the gait recognition result; and
sending, based on identity information of the at least one of the one or more subjects, an unlocking signal to a lock, so that the lock is unlocked.
16. The smart door apparatus of claim 15 , wherein the determining the gait recognition result includes:
determining, based on the one or more gait features using a gait recognition model, a first confidence level that the identity information of the at least one of the one or more subjects corresponds to one of one or more allowed entry identities;
determining the gait recognition by determining whether the first confidence level exceeds or equals a first threshold.
17. The smart door apparatus of claim 16 , wherein the determining the facial recognition result includes:
determining, based on the one or more facial features, a second confidence level that the identity information of the at least one of the one or more subjects corresponds to the one of the one or more allowed entry identities;
determining the facial recognition result by determining whether the second confidence level exceeds or equals a second threshold.
18. The smart door apparatus of claim 17 , wherein the determining, based on the one or more facial features of the at least one of the one or, a second confidence level that the identity information is the one of the one or more allowed entry identities includes:
obtaining a face recognition model; and
determining the second confidence level based on the facial features of the at least one of the one or more subjects using the face recognition model.
19. The smart door apparatus of claim 15 , wherein the method further includes:
sending information associated with the at least one of the one or more subjects to a client terminal based on the identity information;
generating a warning signal for the at least one of the one or more subjects based on the identity information;
sending the information associated with the at least one of the one or more subjects to a client terminal associated with a specific contact based on the identity information;
reporting the police based on the identity information; or
recording the information associated with the at least one of the one or more subjects.
20. A non-transitory computer readable medium storing instructions, the instructions, when executed by at least one processor, causing the at least one processor to implement a method comprising:
obtaining image data of one or more subjects, the image data being acquired by one or more cameras in communication with the computer;
determining, based on the image data, one or more gait features of at least one of the one or more subjects;
determining, based on the gait features of at least one of the one or more subjects, a gait recognition result indicating whether the at least one of the one or more subjects is consistent with one of one or more allowed entry identities;
determining, based on facial features of at least one of the one or more subjects derived from the image data, a facial recognition result indicating whether the at least one of the one or more subjects is consistent with the one of one or more allowed entry identities;
in response to determining that the facial recognition result is different from the gait recognition result, updating gait features or facial features of the one of one or more allowed entry identities based on the facial recognition result and the gait recognition result; and
sending, based on identity information of the at least one of the one or more subjects, an unlocking signal to a lock, so that the lock is unlocked.