On-Chip Neural Imaging Sensor for Secure Region Extraction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional image processing technologies require external processors for complex operations like face authentication, potentially compromising image data security and privacy by outputting unprocessed images externally.

Innovation Solution

A solid-state imaging device with an imager, a processing unit using a neural network calculation model to extract specific regions from image data, and an output unit that fabricates and outputs processed image data within the chip, preventing external exposure of unprocessed data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If complicated processing such as face authentication is executed at an external application processor, then processing capability is improved, but security and privacy are compromised due to output of unprocessed image data

Engineering Contradiction:
Improveprocessing capabilityVSAvoidsecurity and privacy
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent merges the image processing functionality into the image sensor chip by integrating a neural network processing unit that can execute fabrication processing directly on the chip. This combines the image capture function with the complex processing capability in a single integrated system, eliminating the need to output unprocessed image data externally while maintaining high processing capability for tasks like face authentication.

Inventive Principle:
Principle #5Merging (Combining)

2Adaptability or versatility

If fabrication processing is executed externally, then processing flexibility is improved, but data transmission volume and security risks increase

Engineering Contradiction:
Improveprocessing flexibilityVSAvoiddata transmission volume
Core Design Contradiction:
Adaptability or versatilityVSLoss of information

Solution Approach 1:

The patent extracts only the necessary processed information (fabrication results) from the image data while leaving the chip, rather than transmitting the entire unprocessed image data externally. The neural network processing unit performs fabrication processing internally and outputs only the essential results, significantly reducing data transmission volume while maintaining processing flexibility.

Inventive Principle:
Principle #2Taking out (Extraction)

3Device complexity

If all image processing is executed externally, then device complexity is reduced, but security control and data privacy are weakened

Engineering Contradiction:
Improvesystem structureVSAvoidsecurity control
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent segments the processing system into two parts: simple capture functions remain with the image sensor, while complex fabrication processing is handled by the integrated neural network processing unit on the same chip. This segmentation allows the system to maintain simplicity in terms of data flow (no external transmission of raw images) while enhancing security control through on-chip processing capability.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS11820289B2Solid-state imaging device and electronic device
Publication Date: 2023.11.21 SONY SEMICON SOLUTIONS CORP
  • US11820289B2 patent drawing
  • US11820289B2 patent drawing
  • US11820289B2 patent drawing

AI summary

A solid-state imaging device includes: an imager (11) configured to acquire image data; a processing unit (14) configured to execute, on the image data or data based on the image data, processing of extracting a specific region based on a neural network calculation model; and an output unit (16) configured to output image data fabricated based on the specific region or the image data read from the imager based on the specific region.