Questions about how individual parts of claim 1 might be changed. A question may describe something different from what the claim language recites, or a version of it; no view is offered on which.
1.1Image from a cabin camera
What ifWhat if a roof-mounted cabin camera supplied the image for cabin registration?
Claim language“a field-of-view image from the smart glasses”
The vehicle processor matches dashboard and pillar landmarks in the roof camera image to the cabin model. The processor combines the calibrated camera position with glasses pose measurements from a separate head tracker to place the cues in the driver's view.
1.2Pose from pixel intensities
What ifWhat if the system estimated pose by minimizing intensity differences across image regions?
Claim language“performing feature point matching”
The processor projects textured surfaces from the cabin model into the glasses camera view using an initial pose estimate. It adjusts the six pose parameters to minimize pixel intensity residuals over dashboard and pillar regions, and passes the resulting pose to the renderer.
1.3Image rendering on the glasses
What ifWhat if the glasses processor generated the assisted driving image?
Claim language“the in-vehicle system generating an assisted driving image”
The vehicle processor sends cue identifiers, anchor coordinates and a timestamp. The glasses processor combines the coordinates with its latest head pose, draws the corresponding symbols into a frame buffer and sends the frame to the display.
1.4Windshield display for driving cues
What ifWhat if a windshield head-up display presented the assisted driving image?
Claim language“enabling the smart glasses to display the assisted driving image”
The glasses supply head pose to the vehicle processor. The processor prewarps the cue for windshield curvature and the measured viewing position, and a dashboard projector directs the resulting image onto a reflective region of the windshield.
Section 2 · machine-generated
Variations
Further arrangements of the same subject matter, described here as of the publication date at the foot of this page.
Other field of use
Valve guidance at a pumping station
A local controller stores a three-dimensional model of a pumping station with valve locations and receives camera frames from wearable display glasses. The controller matches image features on pipe supports and equipment housings to mapped points and estimates the camera pose. It projects the valve selected in the work order into the camera view, renders an outline at that location, and sends the rendered image to the glasses.
Component substitution
Inertial and optical pose tracking
Four calibrated infrared cameras mounted at the corners of the cabin track six infrared emitters fixed to the glasses in a known three-dimensional pattern. The in-vehicle computer triangulates the emitter positions and fuses the resulting glasses pose with accelerometer and gyroscope readings from the glasses using an extended Kalman filter. It uses the fixed transform from the emitter pattern to the glasses camera to project cabin model coordinates into the camera frame, renders the overlay, and transmits the image to the glasses.
Parameter or scale variation
Dual rate registration and rendering
The glasses capture 1280 by 720 pixel frames at 60 frames per second and sample an inertial sensor at 240 hertz. The in-vehicle processor matches features in every fourth frame to the cabin model, producing a visual pose correction at 15 hertz. It propagates the pose between corrections using the inertial samples, renders an overlay for each camera frame, and sends the rendered images to the glasses at 60 frames per second.
Combination with existing systems
Registered overlays for trailer alignment
A trailer reversing system supplies the measured hitch angle, steering angle, and trailer dimensions to the in-vehicle processor. The processor uses these inputs in a low-speed kinematic model to calculate the paths of the trailer wheels over the next three meters. It matches features in the glasses camera image to the cabin model, projects the calculated paths into the driver's view of the side mirrors using calibrated mirror geometry, and sends the overlay image to the glasses.
Instrument overlays from vehicle diagnostics
An onboard diagnostic gateway supplies fault status and measured values to the in-vehicle processor, which stores a table linking each supported diagnostic signal to a cabin control or instrument. The processor matches features in the glasses image to the cabin model and estimates the camera pose. When a supported fault is active, it projects the associated control or instrument location into the image, renders a label with the diagnostic status and measured value, and sends the image to the glasses.
Automated or data-driven variant
Cabin model updates from parked scans
A depth camera on the glasses records synchronized color and depth frames during a cabin scan while the vehicle is parked. The in-vehicle processor registers successive depth frames against fixed cabin landmarks, merges the depth measurements into a surface model, and stores image descriptors with their three-dimensional locations. During operation, the processor matches features in live glasses frames to those descriptors, estimates the camera pose, and sends the rendered assistance image to the glasses.
Specific version of a part
Live ranging to display anchors
Four ranging transceivers fixed around the instrument panel exchange radio pulses with two transceivers on the glasses frame. An initialization step measures the fixed transceivers' relative positions, and the processor combines subsequent range measurements with inertial data to position a cue beside a selected transceiver.
Passive reflection of dashboard projection
A dashboard projector forms a cue pattern, and relay optics direct its light toward a partially reflective plate in the glasses. A cabin head tracker drives a steering mirror that keeps the beam on the plate as the driver's head moves.
Calibrated mesh and grayscale frames
The vehicle processor loads a textured triangular mesh of the dashboard, windshield surround and pillars in vehicle coordinates. The glasses send timestamped grayscale frames from a forward-facing camera with calibrated focal length and lens distortion.
Binary descriptors for cabin landmarks
The processor extracts image corners and compares binary intensity descriptors with descriptors attached to three-dimensional cabin landmarks. It estimates camera position and orientation from the matched pairs and removes pairs with large reprojection errors.
Depth tested lane guidance
The renderer uses the field-of-view image as a background layer and projects lane guidance geometry from vehicle coordinates using the estimated camera pose. It renders the cabin mesh into a depth buffer and clips cue pixels behind the dashboard or pillars.
Wireless frames with pose timestamps
The vehicle processor compresses the rendered frames and sends them with pose timestamps over a short-range radio link. The glasses decode the frames, apply a rotational correction using the latest gyroscope readings and present the corrected frames on the near-eye display.
Section 3 · quoted from USPTO records
Other work
Passages reproduced word for word from two earlier-published U.S. patent documents, placed beside the language of claim 1. No characterization is offered. Each passage is marked A or B for the document it comes from; ¶ numbers are the paragraph numbers in the USPTO’s published full text of that document.
1.1the in-vehicle system obtaining a vehicle cabin model corresponding to a vehicle operated by the in-vehicle system and a field-of-view image from the smart glasses;
In various embodiments, processor 30 may receive or construct image information 32 corresponding to the component 26. Processor 30 may further include a reference model 34 stored, for example, in memory of processor 30.
A · US 10,958,843 B2 · ¶ 37
1.2the in-vehicle system performing feature point matching between the field-of-view image and the vehicle cabin model to determine spatial position information;
The second camera 44 can be configured, upon a calibration, for an image-to-model registration 60 of a part/component 26 being viewed. The image-to-model registration 60 of the part/component 26 is translated to the first camera via a precomputed extrinsic relationship 62 between the first camera 42 and the second camera 44.
A · US 10,958,843 B2 · ¶ 42
1.3the in-vehicle system generating an assisted driving image based on the field-of-view image, the vehicle cabin model, and the spatial position information; and
1.4the in-vehicle system transmitting the assisted driving image to the smart glasses, and enabling the smart glasses to display the assisted driving image.
User interface 40 may include a display. Optical inspection system 10 may display an indication of the defect to component 26, which may include an image and/or a report.
A · US 10,958,843 B2 · ¶ 38
When such a combined image is outputted as an image to be displayed on the HMD, the combined image is an image close to what can be seen from the HMD.
B · US 2024/0281939 A1 · ¶ 39
An empty cell means only that no passage was reproduced there.
Statements in these documents
What is needed is a way to compute a zoom factor that is not based on structural details of the component being inspected.
A · US 10,958,843 B2 · ¶ 2
Particularly, the disclosure is directed to obtaining an accurate lens zoom factor for the narrow field of view camera by coupling another camera having a wider field of view to move simultaneously with the narrow field of view camera.
A · US 10,958,843 B2 · ¶ 1
In a case of applying image processing to a non-perspective projection image, such as a fisheye image, on the assumption that the image will be viewed in an HMD, a technique in the specification of US-2020-0382755 requires changing the image processing in accordance with the projection method and the field of view in order to achieve a uniform image processing result for all field of views.
B · US 2024/0281939 A1 · ¶ 4
One aspect of the embodiments provides a technique for enabling image processing for perspective projection to be applied to non-perspective projection images independently of the region in the image.
B · US 2024/0281939 A1 · ¶ 5
Cite aspatent.fyi, US 2026/0356196 A1, claim 1, first published 2026-11-19T22:03:07-05:00, https://patent.fyi/US20260356196/, SHA-256 296428bca5dc5cc2a47f48d01f6aeb10bfaa038616937f5656e40fbffadb611a