Motion vector identification in a Wi-Fi motion detection system
Systems and methods for detecting small motion vectors in a Wi-Fi motion detection system are provided. A channel is monitored over time and then evaluated to see if there is a difference, or delta, between the current time sample and other similar time samples. The delta is compared to well-known historical data that is stored in a database and used as a baseline for no motion to identify what the small motion vector is. If the small motion vector activity can't be identified, a learning module is executed to determine the closest match.
1. A method for detecting motion vectors, the method comprising:
receiving indication of an identified activity, the indication based on data collected by a wireless access point that monitors a target environment;
comparing the identified activity to one or more activities stored in a trigger database, wherein at least one of the stored activities include a period of no motion;
determining that the identified activity matches the at least one of the stored activities;
receiving response data regarding a predetermined time period following the identified activity within the target environment;
calculating a delta between the response data and each of one or more sets of historical response data indicative of no motion, the one or more sets stored in a historical sample database; and
identifying that the response data is indicative of a motion activity based on comparing the calculated delta to one or more sets of motion vector activity data stored in a motion vector database.
2. The method of claim 1 , wherein identifying that the response data is indicative of the motion activity is further based on an exact match to at least one of the sets of motion vector activity data in the motion vector database, and further comprising adding the response data to the at least one set of motion vector activity data in the motion vector database.
3. The method of claim 1 , wherein identifying that the response data is indicative of the motion activity is further based on an inexact match to at least one of the sets of motion vector activity data in the motion vector database, the inexact match associated with an identified similarity percentage above a predetermined threshold, and further comprising adding the response data to the at least one set of motion vector activity data in the motion vector database based on the identified similarity percentage.
4. The method of claim 3 , further comprising identifying the inexact match to the at least one set, wherein identifying the inexact match comprises:
calculating a similarity percentage measuring similarity of data points between the response data and each of the sets of motion vector activity data in the motion vector database;
identifying one or more of the similarity percentages that are above the predetermined threshold; and
selecting the at least one set of motion vector activity data based on the identified similarity percentage being the highest similarity percentage among the identified similarity percentages.
5. The method of claim 1 , wherein identifying that the response data is indicative of the motion comprises:
identifying that the response data does not match any of the sets of motion vector activity data in the motion vector database;
prompting a user device to identify a motion vector activity associated with the response data;
receiving an identification for the motion vector activity from the user device; and
adding a new set of motion vector activity data associated with the identified motion vector activity to the motion vector database, the new set including the response data.
6. The method of claim 5 , wherein the new set of motion vector activity data is initially added in association with an unknown activity indicator to the motion vector database, and wherein adding the new set includes updating the unknown activity based on the identification from the user device.
7. The method of claim 1 , wherein the response data includes a location, time, and trigger activity associated with the identified motion activity.
8. A system for detecting motion vectors, the system comprising:
a wireless access point located in a target environment, wherein the wireless access point monitors the target environment; and
a cloud server that:
receives indication of an identified activity, the indication based on data collected by the wireless access point;
compares the identified activity to one or more activities stored in a trigger database, wherein at least one of the stored activities include a period of no motion;
determines that the identified activity matches the at least one of the stored activities;
receives response data regarding a predetermined time period following the identified activity within the target environment;
calculates a delta between the response data and each of one or more sets of historical response data indicative of no motion, the one or more sets stored in a historical sample database; and
identifies that the response data is indicative of a motion activity based on comparing the calculated delta to one or more sets of motion vector activity data stored in a motion vector database.
9. The system of claim 8 , wherein the cloud server identifies that the response data is indicative of the motion activity further based on an exact match to at least one of the sets of motion vector activity data in the motion vector database, and wherein the cloud server further adds the response data to the at least one set of motion vector activity data in the motion vector database.
10. The system of claim 8 , wherein the cloud server identifies that the response data is indicative of the motion activity further based on an inexact match to at least one of the sets of motion vector activity data in the motion vector database, the inexact match associated with an identified similarity percentage above a predetermined threshold, and wherein the cloud server further adds the response data to the at least one set of motion vector activity data in the motion vector database based on the identified similarity percentage.
11. The system of claim 10 , wherein the cloud server further identifies the inexact match to the at least one set, wherein the cloud server identifies the inexact match by:
calculating a similarity percentage measuring similarity of data points between the response data and each of the sets of motion vector activity data in the motion vector database;
identifying one or more of the similarity percentages that are above the predetermined threshold; and
selecting the at least one set of motion vector activity data based on the identified similarity percentage being the highest similarity percentage among the identified similarity percentages.
12. The system of claim 8 , wherein the cloud server identifies that the response data is indicative of the motion activity by:
identifying that the response data does not match any of the sets of motion vector activity data in the motion vector database;
prompting a user device to identify a motion vector activity associated with the response data;
receiving an identification for the motion vector activity from the user device; and
adding a new set of motion vector activity data associated with the identified motion vector activity to the motion vector database, the new set including the response data.
13. The system of claim 12 , wherein the new set of motion vector activity data is initially added in association with an unknown activity indicator to the motion vector database, and wherein the cloud server adds the new set by updating the unknown activity based on the identification from the user device.
14. The system of claim 8 , wherein the response data includes a location, time, and trigger activity associated with the identified motion activity.
15. A non-transitory, computer-readable storage medium, having embodied thereon a program executable by a processor to perform a method for detecting motion vectors, the method comprising:
receiving indication of an identified activity, the indication based on data collected by a wireless access point that monitors a target environment;
comparing the identified activity to one or more activities stored in a trigger database, wherein at least one of the stored activities include a period of no motion;
determining that the identified activity matches the at least one of the stored activities;
receiving response data regarding a predetermined time period following the identified activity within the target environment;
calculating a delta between the response data and each of one or more sets of historical response data indicative of no motion, the one or more sets stored in a historical sample database; and
identifying that the response data is indicative of a motion activity based on comparing the calculated delta to one or more sets of motion vector activity data stored in a motion vector database.