Amazon Technologies, Inc.

Roman Shmulevich

David Bruce McCalib, Jr

Tak Keung Joseph Lui

Oleg Kantor

Alton Paul Werronen

Keith Deacon

Christopher Raymond Grajewski

Animesh Mishra

Joshua David Landry

Vikram Raghavan

Taylor Jeffrey Lloyd

Fabien Lepoutre

Mark G McDonald

Dylan Andrew Norris

Patrick Rubin

A camera-based barcode reader captures an image, determines a region of interest of a barcode within the image, and generates a barcode image from the region of interest. The camera-based barcode reader aligns each of multiple horizontal regions within the barcode image to a reference horizontal region within the barcode image. The alignment of a horizontal region includes (i) determining a set of alignment parameters to apply to the horizontal region based on a comparison of the reference horizontal region to the horizontal region when adjusted to different combinations of alignment parameters and (ii) adjusting the horizontal region according to the set of alignment parameters for that horizontal region. After aligning the barcode image, the camera-based barcode reader decodes the barcode from the barcode image.

Efficient parallel barcode subpixel alignment

Embodiments herein describe design techniques for generating a fingerprint for a shipping label using symbols or bit segments. These symbols or bit segments can be read using lower resolution cameras at greater distances than 1D or 2D barcodes in the shipping labels. In one embodiment, a label designer identifies blank zones in the design of a shipping label and selects symbols to place in those zones. The label designer can select random types of symbols and random sizes for those symbols. When reading the shipping label, a CV system can identify the symbols and their sizes and generate the fingerprint. In another embodiment, the label designer places bit segments in the identified blank zones. The CV system can then read the blank zones in a predefined order and merge the bits segments to form a digital fingerprint for the shipping label.

Package identification using graphics on shipping labels

Embodiments herein describe a package management system that uses two operation modes of RFID readers (i.e., an inventory mode and a search mode) to provide feedback to an associate when a package has been mis-sorted. In one embodiment, the RFID readers are attached to one or more RFID antennas which each are disposed overhead a respective container or storage location in a warehouse. For example, each of the antennas may have a beam pattern or read region that covers the container. Using the RFID antennas, the RFID reader can query the RFID tags in the container to identify the packages stored in the container. In one embodiment, the RFID reader switches to the search mode after a new package is identified. Using the search mode, the RFID reader can quickly identify when the package is stored in the container and provide prompt feedback to the associate.

Overhead RFID antenna

A method of interrogating a plurality of RFID tags comprises interrogating a first RFID tag with a read command comprising a masking value for a predefined portion of a User Memory Bank of the first RFID tag. The predefined portion stores state information for the first RFID tag. The masking value corresponds to a non-quiesced state relative to the read command interrogation. The method further comprises receiving a response from the first RFID tag that indicates that the first RFID tag is in the non-quiesced state. The method further comprises interrogating the first RFID tag with a write command that instructs the first RFID tag to write a predefined value to the predefined portion of the User Memory Bank, wherein writing the predefined value to the predefined portion places the first RFID tag in a quiesced state relative to a subsequent read command interrogation comprising the masking value.

Dynamic RFID tag singulation using state information masking

The embodiments herein measure the tag to noise ratio (TNR) for a tag on an item (e.g., a package) to determine a location of the item in a warehouse. In one embodiment, the TNR is derived by comparing a measured performance parameter for a tag of interest to the combined measured performance parameters for other tags measured by the same reader or for the same tag when measured by a different RFID reader. The higher the TNR, the greater likelihood the tag is within an area assigned to the reader. In another embodiment, the TNR is derived from comparing the signal strength of a tag as measured by multiple RFID readers. In another example, respective TNR values for a plurality of tags detected by a reader can be compared to determine which tag is being carried by an associate.

RFID tag singulation in dense RFID environments

Growing community of inventors

Woodbridge, Canada

Roman Shmulevich