Back to results
US 10,572,963 B1Granted
Detection of items
AssigneeSynapse Technology Corporation
InventorsIan Cinnamon, Bruno Brasil Ferrari Faviero, Simanta Gautam
Life of this patent
Estimating the term…
Abstract
According to an aspect, a system comprises at least one processor, a memory, and a non-transitory computer-readable storage medium storing instructions. The stored instructions are executable to cause the at least one processor to: receive a digital image that represents an object scanned by a detection device, determine a region of the digital image that is likely to contain an item, transform the region of the digital image to an embedding, classify, based on the embedding, the region as containing a known class of known item, and responsive to classifying the region as containing the known class of item: generate a graphical representation based on the known class of item.
Claims20
A method comprising:
- based on a set of training image data, training at least one neural network for identification of items deemed to be of interest, wherein at least one training image data of the set of training image data comprises an item that has been deemed to be of interest;
- receiving, from a detection device of a particular security screening system, image data that represents an object scanned by the detection device;
- based at least on the image data and the at least one neural network:
- identifying one or more regions of interest within the image data, wherein each region of interest has a likelihood of comprising an item of interest that exceeds a threshold likelihood;
- for each given region of interest, generating a vector-space representation corresponding to the given region of interest based on image data within the given region of interest;
- and
- based on one or more vector-space representations and a confidence threshold defined for the particular security screening system that is utilized by the at least one neural network, determining that at least one region of interest of the one or more regions of interest comprises an identified item that has been deemed to be of interest for the particular security screening system;
- and
- based at least on (i) determining that the at least one region of interest comprises the identified item that has been deemed to be of interest for the particular security screening system and (ii) one or more security parameters for the particular security screening system, determining that the identified item is deemed to be a security interest for the particular security screening system;
- based at least on determining that the identified item is deemed to be a security interest for the particular security screening system, presenting, at a display of a computing device of the security screening system, a visualization corresponding to the identified item.
7 more elements
A computing system comprising:
- at least one processor;
- a non-transitory computer-readable medium;
- and
- program instructions stored on the non-transitory computer-readable medium that are executable by the at least one processor to cause the computing system to:
- based on a set of training image data, train at least one neural network for identification of items deemed to be of interest, wherein at least one training image data of the set of training image data comprises an item that has been deemed to be of interest;
- receive, from a detection device of a particular security screening system, image data that represents an object scanned by the detection device;
- based at least on the image data and the at least one neural network:
- identify one or more regions of interest within the image data, wherein each region of interest has a likelihood of comprising an item of interest that exceeds a threshold likelihood;
- for each given region of interest, generate a vector-space representation corresponding to the given region of interest based on image data within the given region of interest;
- and
- based on one or more vector-space representations and a confidence threshold defined for the particular security screening system that is utilized by the at least one neural network, determine that at least one region of interest of the one or more regions of interest comprises an identified item that has been deemed to be of interest for the particular security screening system;
- and
- based at least on (i) determining that the at least one region of interest comprises the identified item that has been deemed to be of interest for the particular security screening system and (ii) one or more security parameters for the particular security screening system, determine that the identified item is deemed to be a security interest for the particular security screening system;
- based at least on determining that the identified item is deemed to be a security interest for the particular security screening system, present, at a display of a computing device of the security screening system, a visualization corresponding to the identified item.
11 more elements
A tangible, non-transitory computer-readable medium comprising program instructions that, when executed by at least one processor, cause a computing system to:
- based on a set of training image data, train at least one neural network for identification of items deemed to be of interest, wherein at least one training image data of the set of training image data comprises an item that has been deemed to be of interest;
- receive, from a detection device of a particular security screening system, image data that represents an object scanned by the detection device;
- based at least on the image data and the at least one neural network:
- identify one or more regions of interest within the image data, wherein each region of interest has a likelihood of comprising an item of interest that exceeds a threshold likelihood;
- for each given region of interest, generate a vector-space representation corresponding to the given region of interest based on image data within the given region of interest;
- and
- based on one or more vector-space representations and a confidence threshold defined for the particular security screening system that is utilized by the at least one neural network, determine that at least one region of interest of the one or more regions of interest comprises an identified item that has been deemed to be of interest for the particular security screening system;
- and
- based at least on (i) determining that the at least one region of interest comprises the identified item that has been deemed to be of interest for the particular security screening system and (ii) one or more security parameters for the particular security screening system, determine that the identified item is deemed to be a security interest for the particular security screening system;
- based at least on determining that the identified item is deemed to be a security interest for the particular security screening system, present, at a display of a computing device of the security screening system, a visualization corresponding to the identified item.
7 more elements
Dependencies are inferred from the claim text.