A monitoring apparatus includes a camera interface, a terminal interface, a communication interface, and a processor. The processor is configured to recognize actions of a settler at a point-of-sale terminal based on images from a camera, recognize input operations performed at the point-of-sale terminal by the settler based on information from the point-of-sale terminal and detect fraudulent acts based on a recognized action and a recognized input operation at the point-of-sale terminal. The processor outputs a fraud detection image corresponding to a detected fraudulent act to at least one of the point-of-sale terminal and an attendant terminal.

According to one embodiment, a commodity registration device includes an optical reading unit with a reading range. The optical reading unit is configured to read a code symbol on a commodity in the reading range. A wireless tag reading unit is configured to read a commodity code from a wireless tag of the commodity in the reading range. A processor determines whether the commodity is in the reading range of the optical reading unit and causes a notification unit to indicate the commodity has been registered if the wireless tag reading unit successfully reads the commodity code from the wireless tag while the commodity is determined to be in the reading range of the optical reading unit.

Examples provide a tag manager component that identifies a plurality of stationary RFID tags located within a three-dimensional space outside an item display area based on an analysis of RFID tag data associated with the plurality of RFID tag readers. A zone detection component analyzes item data associated with a plurality of items corresponding to the plurality of stationary RFID tags and location data associated with the three-dimensional space using a set of per-item criteria. The set of per-item criteria includes a per-item minimum threshold number of items per unit of three-dimensional space and/or a minimum threshold stationary time-period. A verification component analyzes sensor data and/or transaction data to verify whether the three-dimensional space is a termination zone. If the three-dimensional space is a termination zone, an inventory manager component removes the plurality of items from perpetual inventory.

Disclosed are systems and methods including starting with a first number of images, generating a second number of images by digital operations on the first number of images, extracting features from the second number of images, and generating a classification model by training a neural network on the second number of images wherein the classification model provides a percentage likelihood of an image's categorisation, embedding the classification model in a processor and receiving an image for categorisation, wherein the processor is in communication with a POS system, the processor running the classification model to provide output to the POS system of a percentage likelihood of the image's categorisation.

A mobile registration terminal operable by a customer in a store includes a memory, a display, a scanner through which a commodity sold at the store or a medium storing clerk information about a clerk of the store is scanned, and a processor. The processor is configured to, when a commodity is scanned through the scanner, determine whether the commodity is an age-restricted commodity, upon determining that the commodity is not an age-restricted commodity, register the commodity, and upon determining that the commodity is an age-restricted commodity, control the display to display a screen through which an age of the customer is verified by a clerk of the store, and upon receipt of an input of clerk information about a clerk through the scanner, store the clerk information in the memory for checkout of the commodity without further verification of the age of the customer.

A store system includes a server and a store terminal. The server receives data based on purchased commodities respectively input in a plurality of mobile terminals. The server stores the received data based on the purchased commodities. The store terminal acquires, from the stored data, the data based on the purchased commodity input in any one of the mobile terminals. The store terminal instructs correction of the acquired data. The server executes the instructed correction concerning the stored data.

A retail shelf floor divider or shelf floor limiter, which comprises mechanical coupling means which are arranged for coupling the retail shelf floor divider or shelf floor limiter to a further structure, in particular a shelf floor in order to divide the shelf floor in shelf floor sections or to limit it, and electronic detection means for the detection of movement in the neighborhood of the retail shelf floor divider or the shelf floor limiter, which are arranged to utilize an electromagnetic wave, preferred a radio wave, more preferred a light wave, or a sound wave for said detection.

In some embodiments, apparatuses and methods are provided herein useful to self-checkout verification at a retail facility. In some embodiments, there is provided a system for self-checkout verification at a retail facility including a first optical imaging unit; and a control circuit. The control circuit configured to: receive purchase receipt data; receive one or more images of the items in the container; and execute a machine learning model trained to: perform item detection, item classification, and item verification of each item shown in the one or more images; and output electronic data corresponding to an electronic receipt of the items in the container. The control circuit may automatically detect each unpaid item in the container based on a comparison of the purchase receipt data with the electronic data; and provide an alert signal in response to automatically detecting an unpaid item.

A machine-learning algorithm is trained with features relevant to transaction exceptions, distributions of items in transaction mapped to product hierarchies, and operator data. The trained algorithm is trained to predict whether a given transaction requires a transaction exception for potential fraud or for management approval. The trained algorithm is then provided a set of in-progress input data for an in-progress transaction being processed on a transaction terminal. Output from the trained algorithm is used to determine whether the in-progress transaction is allowed to continue processing unabated or whether the in-progress transaction is to be suspended with a transaction exception requiring a manager override or security credential to continue processing.

In some arrangements, product packaging is digitally watermarked over most of its extent to facilitate high-throughput item identification at retail checkouts. Imagery captured by conventional or plenoptic cameras can be processed (e.g., by GPUs) to derive several different perspective-transformed views—further minimizing the need to manually reposition items for identification. Crinkles and other deformations in product packaging can be optically sensed, allowing such surfaces to be virtually flattened to aid identification. Piles of items can be 3D-modelled and virtually segmented into geometric primitives to aid identification, and to discover locations of obscured items. Other data (e.g., including data from sensors in aisles, shelves and carts, and gaze tracking for clues about visual saliency) can be used in assessing identification hypotheses about an item. Logos may be identified and used—or ignored—in product identification. A great variety of other features and arrangements are also detailed.