Compression-Efficient Feature Extraction Method for a CMOS Image Sensor
Open Access
Abstract
To address the power constraints of the emerging Internet of Things (IoT) era, we propose a compression-efficient feature extraction method for a CMOS image sensor that can extract binary feature data. This sensor outputs six-channel binary feature data, comprising three channels of binarized luminance signals and three channels of horizontal edge signals, compressed via a run length encoding (RLE) method. This approach significantly reduces data transmission volume while maintaining image recognition accuracy. The simulation results obtained using a YOLOv7-based model designed for edge GPUs demonstrate that our approach achieves a large object recognition accuracy () of 60.7% on the COCO dataset while reducing the data size by 99.2% relative to conventional 8-bit RGB color images. Furthermore, the image classification results using MobileNetV3 tailored for mobile devices on the Visual Wake Words (VWW) dataset show that our approach reduces data size by 99.0% relative to conventional 8-bit RGB color images and achieves an image classification accuracy of 89.4%. These results are superior to the conventional trade-off between recognition accuracy and data size, thereby enabling the realization of low-power image recognition systems.
Keywords
Funding Information
- JSPS KAKENHI (JP24K15001)
This publication has 23 references indexed in Scilit:
- [Paper] Lightweight Object Detection Model for a CMOS Image Sensor with Binary Feature ExtractionITE Transactions on Media Technology and Applications, 2026
- A 0.8- μ m 32-Mpixel Always-On CMOS Image Sensor With Windmill-Pattern Edge Extraction and On-Chip DNNIEEE Solid-State Circuits Letters, 2025
- The Design of a Computer Vision Sensor Based on a Low-Power Edge Detection CircuitSensors, 2025
- On-Chip Data Reduction and Object Detection for a Feature-Extractable CMOS Image SensorElectronics, 2024
- The Open Images Dataset V4International Journal of Computer Vision, 2020
- A Data-Compressive 1.5/2.75-bit Log-Gradient QVGA Image Sensor With Multi-Scale Readout for Always-On Object DetectionIEEE Journal of Solid-State Circuits, 2019
- Visual Wake Words DatasetarXiv, 2019
- An Over 90 dB Intra-Scene Single-Exposure Dynamic Range CMOS Image Sensor Using a 3.0 μm Triple-Gain Pixel Fabricated in a Standard BSI ProcessSensors, 2018
- ImageNet classification with deep convolutional neural networksCommunications of the ACM, 2017
- The Pascal Visual Object Classes (VOC) ChallengeInternational Journal of Computer Vision, 2009