Visual and geolocation analytic system and method

1. A visual and geolocation analytic system including:

an analytic device;

a number of image capturing devices connected to said analytic device;

wherein said image capturing devices capture images of an object at a time interval and send said captured images to said analytic device; said analytic device comprises a deep learning model for analyzing said captured images, allowing said object to be identified and tagged, and allowing a path of movement of said object across time to be tracked;

wherein said deep learning model is trained with image classification using convolutional neural networks and fine tuned with bounding boxes and image distortions, so that it is capable of object detection by recognizing distinct attributes within images and to tag positional information on said object.

2. The visual and geolocation analytic system according to claim 1 , wherein said deep learning model of said analytic device comprises an object detection system for recognizing a stock keep- ing unit of a product item.

3. The visual and geolocation analytic system according to claim 1 , further comprising a transaction device linked to said analytic device, wherein said transaction device sends a transaction record involving said object to said analytic device, said analytic device integrates said transaction record into said path of movement of said object.

4. The visual and geolocation analytic system according to claim 1 , further comprising at least one weight sensor linked to said analytic device, wherein said weight sensor sends a signal to said analytic device when a weight change is detected at a sensed location.

5. The visual and geolocation analytic system according to claim 1 , wherein said captured images are discarded from said analytic device after analyzing, and only a structured data set containing an identity and said path of movement of said object across time is retained.

6. The visual and geolocation analytic system according to claim 1 , wherein said image capturing devices include fish eye cameras.

7. The visual and geolocation analytic system according to claim 1 , wherein said analytic device is synchronized with analytic devices of other systems to form a distributed edge network.

8. The visual and geolocation analytic system according to claim 1 , wherein said deep learning model of said analytic device comprises a facial recognition system for recognizing a person.

9. The visual and geolocation analytic system according to claim 8 , wherein said facial recognition system further determines demographic information of said person.

10. A method for visual and geolocation analysis, comprising the steps of:

capturing images of an object at a time interval with a number of image capturing devices;

recognizing and tagging said object from said captured images by an analytic device having a deep learning model;

extracting time and location data of said object from said captured images, and tracking a path of movement of said object across time;

wherein said deep learning model is trained with image classification using convolutional neural networks and fine tuned with bounding boxes and image distortions, so that it is capable of object detection by recognizing distinct attributes within images and to tag positional information on said object.

11. The method for visual and geolocation analysis according to claim 10 , further comprising the steps of:

receiving a transaction record involving said object from a transaction device; and

integrating said transaction record into said path of movement of said object by said analytic device.

12. The method for visual and geolocation analysis according to claim 10 , further comprising the step of receiving a signal from a weight sensor when a weight change is detected at a sensed location.

13. The method for visual and geolocation analysis according to claim 10 , further comprising the step of discarding said captured images after analyzing.

14. The method for visual and geolocation analysis according to claim 10 , further comprising the step of synchronizing with other analytic devices to form a distributed edge network.

15. A visual and geolocation analytic system including:

an analytic device;

a number of image capturing devices connected to said analytic device;

wherein said image capturing devices capture images of an object at a time interval and send said captured images to said analytic device; said analytic device comprises a deep learning model for analyzing said captured images, allowing said object to be identified and tagged, and allowing a path of movement of said object across time to be tracked;

wherein said deep learning model is trained with image classification using convolutional neural networks and fine tuned with bounding boxes and image distortions, so that it is capable of object detection by recognizing distinct attributes within images and to tag positional information on said object;

wherein said captured images are discarded from said analytic device after analyzing, and only a structured data set containing an identity and said path of movement of said object across time is retained.