Overhead Image Synthesis for Vehicle Display and Recognition Load Reduction

Solution Overview

Problem

Existing systems face increased processing loads due to image geometric transformation processing for multiple applications of overhead images in vehicles, such as combine harvesters, which are not optimized for different uses like display and sensing.

Innovation Solution

An information processing device that includes a viewpoint transformation unit and an overhead image generating unit, which performs different synthesis processing for display and recognition applications, reducing the load by optimizing image transformation for specific uses.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the overhead image is used for multiple applications (display and sensing), then the versatility is improved, but the processing load increases due to repeated image geometric transformation processing

Engineering Contradiction:

Improveapplication versatilityVSAvoidprocessing load

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent segments the synthesis processing into different types (first synthesis processing for display, second synthesis processing for sensing) based on the application purpose. This allows the system to perform only the necessary processing for each application rather than repeatedly performing all transformations, thereby reducing processing load while maintaining versatility.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies different synthesis processing methods tailored to specific local needs: gradient blending for display applications to ensure visual quality, and simple blending for sensing applications to reduce processing load. This localized optimization resolves the contradiction by matching processing intensity to application requirements.

Inventive Principle:

Principle #3Local quality

2Adaptability or versatility

If image geometric transformation processing is performed for multiple applications, then the adaptability is improved, but the processing time increases

Engineering Contradiction:

Improveapplication versatilityVSAvoidprocessing time

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent segments the synthesis processing into different types (first synthesis processing for display, second synthesis processing for sensing) based on the application purpose. This allows the system to perform only the necessary processing for each application rather than repeatedly performing all transformations, thereby reducing processing load while maintaining versatility.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies partial processing actions appropriate to each application: full gradient blending processing for display (where visual quality is critical) and simplified simple blending for sensing (where processing time is less critical). This selective processing approach reduces overall processing time while maintaining adaptability.

Inventive Principle:

Principle #16Partial or excessive action

3Device complexity

If the same synthesis processing is used for all applications, then the device complexity is reduced, but the processing efficiency decreases

Engineering Contradiction:

Improveprocessing complexityVSAvoidprocessing efficiency

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The patent applies different synthesis processing methods tailored to specific local needs: gradient blending for display applications to ensure visual quality, and simple blending for sensing applications to reduce processing load. This localized optimization resolves the contradiction by matching processing intensity to application requirements.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent implements dynamic selection of synthesis processing methods based on the application type. The system can switch between first synthesis processing (gradient blending) and second synthesis processing (simple blending) depending on whether the output is for display or sensing, optimizing efficiency without requiring a completely different processing pipeline for each application.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS20260051180A1Information processing device and information processing method

Publication Date: 2026.02.19 SONY SEMICON SOLUTIONS CORP

AI summary

An information processing device includes: a viewpoint transformation unit that performs viewpoint transformation on a plurality of images generated by a plurality of imaging units mounted on front, rear, left, and right sides of a vehicle; and an overhead image generating unit that generates an overhead image by combining the plurality of images having been subjected to the viewpoint transformation by the viewpoint transformation unit, in which the overhead image generating unit: in a case where a first overhead image to be displayed in the vehicle is generated, executes first synthesis processing; and, in a case where a second overhead image for executing recognition processing for surroundings of the vehicle is generated, executes second synthesis processing different from the first synthesis processing.