Contextual self-checkout based verification

A method, comprising:

• providing executable instructions to a processor of a server causing the processor to execute the executable instructions and perform operations comprising:
• capturing images of a Self-Checkout (SCO) area having a Self-Service Terminal (SST), a bag scale, a produce scale, and shelves;
• identifying from the images a consumer in the SCO area;
• tracking items being processed at the SST from the images during a transaction;
• maintaining a context from the images between each item with respect to:
• the consumer, bags of the consumer, a cart of the consumer, the bag scale, the produce scale, and the shelves, wherein maintaining further includes maintaining the context as item locations for the items within the SCO area, bag locations for bags within the SCO area, cart location for a cart within the SCO area, produce scale location for the produce scale within the SCO area, bag scale location for the bag scale within the SCO area, shelf locations for the shelves within the SCO area, consumer-possessed items detectable within the SCO area, and any individuals present with the consumer in the SCO area, wherein maintaining further includes maintaining analytics with the context, wherein the analytics include a total number of items detected and direction of movement of the consumer away from the terminal;
• receiving an attendant intervention event raised by the SST for a particular item based on an unexpected weight provided by the bag scale or the produce scale;
• and
• using the context to instruct the SST to override the attendant intervention event and continue with a next item of the transaction.

A system, comprising:

• a processor;
• a memory;
• a non-transitory computer-readable storage medium having executable instructions;
• the executable instructions when provided to the processor from the non-transitory computer-readable storage medium into the memory cause the processor to perform operations comprising:
• independently monitoring a transaction area through images taken of the transaction area independent of a transaction terminal that processes a transaction, wherein the independently monitoring further includes monitoring a consumer identified in the images performing the transaction at the transaction terminal with respect to a bag scale of the transaction terminal, a cart of the consumer, one or more bags of the consumer, a produce scale, and shelves identified in the transaction area through the images, wherein independently monitoring further includes maintaining a context for the transaction as item locations for the items within the transaction area, bag locations for the bags within the transaction area, cart location for the cart within the transaction area, produce scale location for the produce scale within the transaction area, bag scale location for the bag scale within the transaction area, shelf locations for the shelves within the transaction area, consumer-possessed items detectable within the transaction area, and any individuals present with the consumer in the checkout area, wherein maintaining further includes maintaining analytics with the context, wherein the analytics include a total number of items detected and direction of movement of the consumer away from the transaction terminal;
• clearing attendant intervention alerts when the context from the images and a transaction state at a transaction terminal of the transaction area indicates the attendant intervention alerts are unnecessary;
• maintaining additional analytics for items processed during a transaction by the transaction terminal;
• comparing the additional analytics, the context, and the analytics for the context to transaction totals when the transaction concludes;
• and
• processing custom-defined actions based on rules associated with the context, the analytics, the additional analytics, and the transaction totals.