System and method for machine learning-driven object detection

1. A gaming monitoring system comprising:

at least one camera configured to capture images of a gaming surface; and

a computing apparatus in communication with the at least one camera, said computing apparatus configured to analyse the captured images of the gaming surface to automatically apply machine learning processes to identify game objects, game events and players in the captured images, wherein

the game objects comprise one or more stacks of one or more wager objects, and

the computing apparatus is further configured to automatically identify and estimate a value of each stack of one or more wager objects by,

identifying one or more first regions of interest in the captured image that relate to one game object using a trained first region proposal network;

identifying at least one subset of first regions of interest among the one or more first regions of interest that relate to a single stack of one or more wager objects using a trained first object detection network;

identifying one or more second regions of interest that relate to part of an edge pattern on each wager object that forms part of the single stack of one or more wager objects in each of the identified at least one subset of first regions of interest using a trained second region proposal network;

identifying a value pattern in each of the one or more second regions of interest using a trained second object detection network; and

estimating a total wager value of the single stack of one or more wager objects in each of the at least one subset of first regions of interest using the identified value patterns and a lookup table.

2. The system of claim 1 , wherein the machine learning processes are implemented through one or more neural networks.

3. The system of claim 2 , wherein the one or more neural networks comprise one or more deep neural networks.

4. The system of claim 3 , wherein the one or more deep neural networks comprise one or more convolutional neural networks.

5. The system of claim 4 , wherein at least one of the one or more convolutional neural networks comprises at least one of: a region proposal network; or an object detection network.

6. The system of claim 1 , wherein the game objects further comprise playing cards, position markers or monetary objects.

7. The system of claim 1 , wherein the said computing apparatus is further configured to associate each of the one or more first regions of interest with a wager area identifier.

8. The system of claim 1 , wherein the computing apparatus is further configured to identify a start and end of a game based on a game start and end trigger configuration stored in a data store accessible to the computing apparatus.

9. The gaming monitoring system of claim 1 , wherein the computing apparatus is further configured to determine an illumination of an indicator light on a dealing device on the gaming surface.

10. A computer system or apparatus configured to execute a neural network system for game object identification, comprising:

at least one processor; and

a memory accessible to the at least one processor and storing code to execute,

a wager object region proposal network (RPN) to receive image data from captured images of a gaming table; and

a wager object detection network to receive an output of the wager object RPN;

wherein the wager object detection network detects one or more wager objects in the captured images based on an output of the wager object region proposal network.

11. The computer system or apparatus of claim 10 , further comprising:

a gaming table region proposal network (RPN) to receive image data from captured images of a gaming table; and

a gaming table object detection network to receive an output of the gaming table RPN;

wherein the gaming table object detection network detects one or more gaming objects in the captured images based on an output of the gaming table object detection network, wherein the one or more gaming objects are different from the one or more wager objects.

12. The system of claim 11 , wherein the wager object region proposal network (RPN) is further configured to identify one or more first regions of interest in the captured image that relate to the one or more wager objects.

13. The system of claim 12 , wherein the wager object region proposal network (RPN) is further configured to determine a wager area identifier associated with the one or more first regions of interest.

14. A gaming monitoring system comprising:

at least one camera configured to capture images of a gaming surface; and

a computing apparatus in communication with the at least one camera, said computing apparatus configured to analyse the captured images of the gaming surface to automatically apply machine learning processes to identify game objects, game events and players in the captured images;

wherein the machine learning processes are implemented through one or more convolutional neural networks, and

wherein the one or more convolutional neural networks comprise a first convolutional neural network for performing image segmentation to determine an outline of a game object in the captured image, wherein the one or more convolutional neural networks comprises a game object classifier neural network configured to classify the game object in the determined outline.

15. The system of claim 14 , wherein

the game objects include wager objects; and

the one or more convolutional neural networks comprise a second convolutional neural network for performing image processing operations to determine a wager value of the wager objects.

16. The system of claim 14 , wherein

the game objects include wager objects; and

the computing apparatus is further configured to determine a wager area identifier based on the determined outline of the wager objects.