Aug 25, 2026

US12716685 - Techniques for determining a grip layout to improve gun ergonomics

The present disclosure provides systems and techniques for determining a grip layout for a gun. The techniques include obtaining an image depicting a hand, obtaining calibration data associated with the image, calculating a real distance between a first landmark of the hand and a second landmark of the hand, and determining a grip layout for the gun based on the real distance between the first landmark and the second landmark. Calibration data may include a scale for the image or data that may be used to derive the scale, such as an object of known dimensions, an angular field of view, or a distance between the device used to capture the image and the hand. The grip layout may include a backstrap size, a sensor size, a sensor location on the gun, or any combination thereof.

Assignee: Biofire Technologies Inc.
Inventors: Bryan Edward Rogers, Jules Woolf Kramer, Benjamin William Dwyer, Shayna Rochelle-Ann Thomson, William Ward, Kai Thorin Kloepfer
Filing date: 2025-08-14

gunergonomicsgriplayout

The patent describes a method for determining an ergonomic grip layout for firearms by analyzing images of a user’s hand using machine learning techniques. This process involves measuring hand dimensions, predicting handedness, and selecting appropriate features such as sensor location and backstrap size to enhance user comfort and control.

Claim 1

1 . A method of determining a grip layout for a gun, the method comprising: obtaining an image depicting a hand and a fiducial marker; performing object detection on the image to identify the hand and the fiducial marker within the image; performing annotation on the hand based on a machine learning model that has been trained with a first set of training data including multiple images containing hands, wherein each image of the multiple images of the first set of training data is annotated; generating a measurement of the hand based on a second machine learning model that has been trained with a second set of training data including multiple images containing hands, wherein each image of the multiple images of the second set of training data is labeled with a measurement; generating a handedness prediction based on a third machine learning model that has been trained with a third set of training data including multiple images containing hands, wherein each image of the multiple images of the third set of training data is labeled as a left hand or a right hand; and selecting (i) a sensor location for the gun based on the handedness prediction and (ii) a backstrap size for the gun based on the measurement of the hand. obtaining an image depicting a hand and a fiducial marker; performing object detection on the image to identify the hand and the fiducial marker within the image; performing annotation on the hand based on a machine learning model that has been trained with a first set of training data including multiple images containing hands, wherein each image of the multiple images of the first set of training data is annotated; generating a measurement of the hand based on a second machine learning model that has been trained with a second set of training data including multiple images containing hands, wherein each image of the multiple images of the second set of training data is labeled with a measurement; generating a handedness prediction based on a third machine learning model that has been trained with a third set of training data including multiple images containing hands, wherein each image of the multiple images of the third set of training data is labeled as a left hand or a right hand; and selecting (i) a sensor location for the gun based on the handedness prediction and (ii) a backstrap size for the gun based on the measurement of the hand.

Google Patents

https://patents.google.com/patent/US12716685

USPTO PDF

https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/12716685