Back to results
US 10,650,232 B2Granted
Produce and non-produce verification using hybrid scanner
AssigneeNCR Corporation
InventorsChao He, Meng Yu, Richard C. Quipanes, Em Parac, Kristoffer Dominic Amora
Life of this patent
Estimating the term…
Abstract
Approaches to produce verification are addressed which may be suitably employed in conjunction with a low resolution image scanner. Features, such as lines, corners, text, texture, edges and the like are extracted and compared with features for an identified item or evaluated to determine if the features collectively are indicative of a produce item or a non-produce item.
Claims20
Apparatus for verifying whether an item is produce or not comprising:
- a scale configured to weigh produce items;
- a hybrid scanner including:
- i) a laser scanner configured to read optical codes affixed to the items and 2) a camera configured to generate an image of the item when the item is placed within a field of view of the camera on the scale, wherein the hybrid scanner is configured to read the optical codes and generate the image at the same time;
- an illumination mechanism configured to selectively illuminate the field of view of the camera;
- a produce verification model database configured to store data for produce items to be verified;
- and
- a processor that executes software, which causes the processor to:
- control the camera and the illumination mechanism;
- retrieve a model for a particular item of produce to be weighed from the produce verification model database;
- control the camera to capture image data for the item;
- extract features from the captured image data;
- use the features to classify the image into categories representative of a produce item or a non-produce item based on:
- line features, corner features, edge features, texture features, and any characters or text features;
- wherein a total number of the line features is a factor relevant to determining the non-produce item as compared to the produce item, wherein a total number of the corner features is a second factor relevant to determining the non-produce item as compared to the produce item, wherein a total number of uniform patterns within the texture features is a third factor relevant to determining the produce item as compared to the non-produce item, and wherein a total number of the characters or the text features is fourth factor relevant to determining the non-produce item as compared to the produce item;
- use the classified image to sub-classify the image from the categories into sub-categories based on particular extracted features representing 1) packaging materials for the item that distinguishes between boxes, bottles, and cans, 2) a design of the packaging materials that distinguishes between high-contrast or low-contrast present in the image, 3) item object shape, 4) item object amount indicating a single item or a bundle of multiple items;
- fit the extracted features, the categories, and sub-categories to a retrieved model for the particular item of produce;
- and verify the item as produce or not based upon the fit of the extracted features to the retrieved model.
14 more elements
Apparatus for categorizing an item as produce or non-produce comprising:
- a hybrid scanner including:
- i) a laser scanner configured to read optical codes affixed to the item, and ii) a camera configured to generate an image of the item when the item is placed within a field of view of the camera on the scale, and wherein the hybrid scanner is configured to read the optical codes and generate the image at the same time;
- a categorization database configured to store data for produce and non-produce items to be categorized;
- and
- a processor configured to execute software, which causes the processor to:
- retrieve product categorization criteria from the categorization database;
- control the camera to capture image data for the item;
- extract features from the captured image data including line features, corner features, edge features, texture features, and any characters or text features;
- apply a retrieved categorization criteria;
- and
- categorize the item as produce or non-produce based upon the application of the categorization criteria using the extracted features from the captured image data;
- wherein a total number of the line features is a first factor relevant to determining the non-produce as compared to the produce, wherein a total number of the corner features is a second factor relevant to determining of the non-produce as compared to the produce item, wherein a total number of uniform patterns within the texture features is third factor relevant to determining the produce as compared to non-produce, and wherein a total number any of the characters or the text features is a fourth factor relevant to the non-produce as compared to the produce;
- wherein classify includes processing to further sub-classify the item based the extracted features into sub-categories identified from:
- 1) packaging materials with the item that distinguishes between boxes, bottles, and cans, 2) a package design for the packaging materials that distinguishes between high-contrast or low-contrast present in the image, 3) item object shape, and 4) item object amount representing a single item or a bundle of multiple items.
11 more elements
Dependencies are inferred from the claim text.