Automatic region of interest detection for casino tables
A method, system, and computer readable storage for identifying regions of interest on a casino gaming table. Regions of interest are locations on an image where relevant actions are typically going to occur. For example, locations of betting area region of interest and card area regions of interest are determined so that these areas can be analyzed on a video camera. The analyzed areas can then be used for things such as determining and tracking player betting amounts, determining if the game is being dealt properly (by analyzing the cards dealt to determine if the proper cards are dealt and/or the proper payouts are made).
1. A method for identifying a region of interest on a casino table, the method, comprising:
defining template meta-data for an image of the casino table, wherein the template meta-data that defines a regions of interest on the template image of the casino table where activity occurs on the casino table;
capturing with a camera a table image of the casino table, wherein a gaming object is placed on the casino table in one of the regions of interest;
generating camera meta-data for the region of interest based upon the location of the gaming object;
defining a homography between the template meta-data and the camera meta-data for the region of interest based on the location of the gaming object in the region of interest; and
generating camera meta-data for a different region of interest based upon the homography.
2. The method as recited in claim 1 , further comprising:
distributing the template meta-data to a casino.
3. The method as recited in claim 1 , wherein the template meta-data is distributed to a casino server at a casino subsequent to the template meta-data being updated.
4. The method as recited in claim 3 , further comprising:
distributing the template meta-data from the casino server to a table computer at a casino table.
5. The method as recited in claim 1 , further comprising:
capturing an image of the casino table with no gaming object present on it;
capturing an image of the casino table with the gaming object present on it; and
subtracting the image of the casino table with no gaming object present on it from the image of the casino table with gaming object present on it.
6. The method as recited in claim 5 , further comprising receiving approval from a technician that the identification of the gaming object on the casino table is identified correctly.
7. The method as recited in claim 1 , wherein the region of interest comprises any of a betting region of interest defining a betting area areas and a card area region of interest defining an area where cards are dealt.
8. The method as recited in claim 1 , further comprising defining template meta-data for a plurality of different regions of interest, wherein different types of regions of interest on the template image are identified by different colors.
9. The method as recited in claim 1 , further comprising defining template meta-data for a plurality of different regions of interest, wherein different regions of interest of a same type on the template image are identified by different colors.
10. An apparatus comprising an electronic processor connected to at least one non-transitory computer readable storage medium having instructions stored thereon which when executed by a processor cause the processor to:
define template meta-data for an image of the casino table, wherein the template meta-data that defines regions of interest on a template image of a casino table where activity occurs on the casino table;
capture with a camera a table image of the casino table, wherein objects are placed on the casino table, wherein objects are not placed in a vacant region of interest;
generate camera meta-data locating region of interest on the at least one table image;
define a homography between the template meta-data and the camera meta-data for the regions of interest based on the placement of the objects on the casino table; and
generate camera meta-data for the vacant region of interest based upon the homography.
11. The apparatus as recited in claim 10 , wherein the computer readable instructions are further programmed to cause the at least one electronic processor to:
distribute the template meta-data to a casino.
12. The apparatus as recited in claim 11 , wherein the computer readable instructions are further programmed to cause the at least one electronic processor to distribute the template meta-data to a casino server at a casino subsequent to the template meta-data being updated.
13. The apparatus as recited in claim 12 , wherein the computer readable instructions are further programmed to cause the at least one electronic processor to:
distribute the meta-data from the casino server to a table computer at a casino table.
14. The apparatus as recited in claim 10 , wherein the computer readable instructions are further programmed to cause the at least one electronic processor to:
capture an image of the casino table with no objects present on it;
a capture an image of the casino table with objects present on it; and
subtract the image of the casino table with no objects present on it from the image of the casino table with objects present on it.
15. The apparatus as recited in claim 14 , wherein the computer readable instructions are further programmed to cause the at least one electronic processor to receive approval from a technician that the identifications of the objects on the casino table are identified correctly.
16. The apparatus as recited in claim 10 , wherein the computer readable instructions are further programmed such that the regions of interest comprise betting regions of interest defining betting areas and card area regions of interest defining areas where cards are dealt.
17. The apparatus as recited in claim 10 , wherein the computer readable instructions are further programmed such that different types of regions of interest on the template image are identified by different colors.
18. The apparatus as recited in claim 10 , wherein the computer readable instructions are further programmed such that different regions of interest of a same type on the template image are identified by different colors.