Character And Pattern Recognition TRIZ Cases

Bayesian track confidence estimation weights sensor false positives, false negatives, and occlusion to improve fused object detection.

A single multi-band transceiver maps missing sub-carrier frequencies to cabin positions for lower-cost, privacy-safe occupant detection.

Ground-plane feature selection and sliding-window normal estimation cut computation while keeping surround-view alignment accurate in vehicles.

A camera compares platform patterns with stored reference images to correct carriage floor height without complex sensor systems.

A multi-surface cockpit sensor layout improves camera and sensing coverage while simplifying hardware, lowering power use, and supporting privacy mode.

Corrected lateral acceleration subtracts road cant effects, avoiding unnecessary running control cancellation on curves and reducing delay.

A high-level goal planner and goal-conditioned offline RL policy split long-horizon control into reachable sub-goals, improving shift robustness.

Real-time curve, road surface, and nearby vehicle data are combined to set safer target speeds and cut collision risk on curved roads.

A motorized sliding camera extends for cabin monitoring and retracts into the host to improve privacy and keep the vehicle interior clean.

Physiological sensing establishes an occupant baseline, then adjusts speed, braking, and acceleration in real time to reduce stress and anxiety.

Image recognition detects the trailer coupler and triggers the bird's-eye view at the right distance for steadier hitch ball alignment.

Structured simulator stages with cognitive tests and physiological sensing reduce bias in driver state data for ADAS model training.

Structured simulator driving with DRTs and multimodal sensing reduces bias in driver drowsiness and cognitive load data for ADAS.

Structured simulator stages with N-back tasks and multimodal sensing reduce bias in driver state data for ADAS model training.

Monotonous simulator driving with visual DRTs and multi-sensor monitoring yields reliable driver drowsiness and cognitive-load data for ADAS.

Non-overlapping pixel groups level detection intensity in ghost imaging, cutting correlation workload while preserving image quality.

Sensor-driven predictive slowdown lowers an autonomous vehicle's speed limit in complex traffic to reduce accident risk without constant delays.

Folded windshield IR imaging captures clear front-row occupant views while reducing steering-wheel occlusion and ambient light interference.

Driving-load sensing adjusts in-vehicle content type and suggestion frequency to improve engagement while limiting driver distraction.

Biometric posture data and BP neural prediction improve seat-position matching for different drivers, boosting comfort and driving safety.

Camera and LIDAR data are fused to interpolate lane markers hidden by the hood, improving front wheel position estimation for driver assistance.

Combining camera images with terminal operation data helps distinguish device use from drowsiness and improves takeover readiness judgment.

Real-time sensor fusion links driver distraction with incident-prone locations to trigger timely in-vehicle warnings and mitigation suggestions.

Maintain awareness of objects hidden from sensors by estimating their location in mapped occlusion zones until visibility returns.

Using mobile, GPS, visual, and sensor tracking, the vehicle can park near or pick up a person without limiting their movement.

A transparent windshield display and gaze-based dimming of a second screen reduce unintended line-of-sight shifts and driver distraction.

Dynamic lane-line selection compares map and detected road markings to preserve horizontal vehicle positioning when vertical estimates drift.

By detecting speed bumps in camera data, the controller holds steering torque through the bump to avoid false lane recognition and lane drift.

Driver head or gaze direction triggers camera switching to keep rearview objects visible when composite views would hide them.

By correcting lane-line distances with stored lane width, this case cuts BSD and LCA false warnings and better identifies lane changes.

Rail-mounted seats, sensors, and actuators reconfigure cabin layouts safely around obstructions for comfort, social use, child care, or cargo.

GNN-based sampling predicts realistic, non-overlapping object trajectories in congested driving scenes while reducing computation.

Driver override events are used to recommend when specific driving assistance functions should be activated or adjusted to match individual preferences.

Combining vital signs, external injury sensing, and crash data helps vehicles rank post-crash life-criticality for faster rescue response.

A DNN uses image-only input plus sensor-derived ground truth to estimate object distance and free-space boundaries accurately on curved roads.

When ADAS control reliability drops, a forward camera backup takes over actuation commands to maintain stable vehicle control during sensor or controller faults.

Camera-based gaze detection blocks in-vehicle control inputs and shows warning images when driver distraction persists.

In-cabin and external sensors capture special ride moments automatically, turning autonomous trips into shareable highlight reels.

Criticality-based hardware timers limit how long vehicle data stays accessible in memory, reducing exposure and freeing storage after expiry.

Sensor-based monitoring detects inappropriate language or movements in vehicles and triggers alerts or incident reports for faster response.

Interference indication in radar and LiDAR detection regions enables proactive alarms and more reliable results under background light noise.

Voice and facial signal analysis triggers an electronic interlock to isolate power or immobilize controls when operator impairment is detected.

Exterior cameras monitor user-defined zones around a parked vehicle, with autonomous repositioning to close blind spots and detect target activity.

A rear vehicle camera uses dual-magnification optics and centered mounting to widen coverage, preserve detail, and reduce image processing lag.

Projected light marks on the trailer let a rear camera measure hitch angle accurately without trailer-specific hardware or markings.

A digital twin trained on driver gaze and vehicle behavior predicts hazard perception impairment in real driving and supports timely vehicle intervention.

Sensor-based door hold control keeps a vehicle door closed at the stop when a passenger's item is still detected in the cabin.

Removes projected road notification patterns from vehicle camera images to avoid confusion with parking lines and support accurate parking.