Systems and methods for merchandise automatic checkout
Systems and methods for automatically checking out items located on a moving conveyor belt for the purpose of increasing the efficiency of a checkout process and revenue at a point-of-sale. The system includes a conveyor subsystem for moving the items, a housing that enclosed a portion of the conveyor subsystem, a lighting subsystem that illuminates an area within the housing, visual sensors that can take images of the items including UPCs, and a checkout system that receives the images from the visual sensors and automatically identifies the items. The system may include a scale subsystem located under the conveyor subsystem to measure the weights of the items, where the weight of each item is used to check if the corresponding item is identified correctly. The system relies on matching visual features from images stored in a database to match against features extracted from images taken by the visual sensors.
1. A system for automatically checking out an object, comprising:
a scale subsystem for measuring a weight of the object;
a visual sensor for capturing an image of the object;
a visual subsystem coupled to the visual sensor, wherein the visual subsystem is configured to:
i) extract multiple scale-invariant visual features from the image, each of the multiple scale-invariant visual features having an associated feature descriptor defined by a feature vector;
ii) compare the extracted feature vectors to a plurality of feature vectors associated with a plurality of known objects, each of the known objects having multiple ones of the plurality of feature vectors associated therewith, and
iii) identify multiple ones of the extracted feature vectors that match a subset of the feature vectors associated with one of the known objects; and
a check out subsystem, coupled to the visual subsystem and the scale subsystem, configured to identify the object based on the matching feature vectors and the measured weight.
2. The system of claim 1 , further comprising:
a conveyor subsystem located adjacent to the scale subsystem configured to move the object.
3. The system of claim 2 , further comprising:
a housing that encloses a portion of the conveyor subsystem.
4. The system of claim 3 , wherein the visual sensor is mounted on the housing.
5. The system of claim 4 , further comprising:
a lighting subsystem that illuminates an area within the housing and is mounted on the housing.
6. The system of claim 5 , further comprising:
a motion trigger subsystem for detecting the object and triggering the lighting subsystem and the visual sensor.
7. The system of claim 1 , wherein the visual subsystem is a server.
8. The system of claim 1 , wherein a color and a shape of the object are used to identify the object.
9. The system of claim 1 , wherein, upon identification of the object, the checkout subsystem retrieves color and shape information from a database, and compares the retrieved color and shape information with color and shape information from the image to verify that the identification is correct.
10. The system of claim 1 , wherein the visual subsystem is further configured to determine a size of the object.
11. The system of claim 1 , wherein the visual sensor comprises a conventional UPC reader for reading a UPC on the object.
12. The system of claim 1 , further comprising:
a UPC reader subsystem configured to read a UPC on the object depicted in the image.
13. The system of claim 1 , wherein the checkout subsystem retrieves weight information associated with the identified object upon recognition of the object, and compares the measured weight of the object with the retrieved weight information to verify that the recognition is correct.
14. The system of claim 1 , wherein the image from which the visual subsystem is configured to extract the multiple scale-invariant visual features is raw image data from the visual sensor.