Design of an Edge-Detection CMOS Image Sensor with Built-in Mask Circuits

Area of Science:

• Computer Vision
• Semiconductor Devices
• Image Sensors

Background:

• Traditional computer vision systems often require complex image signal processing for edge detection.
• Low-power consumption is a critical requirement for many embedded and mobile vision applications.

Discussion:

• The proposed CMOS image sensor (CIS) integrates mask circuits directly onto the sensor for selective edge-detection or normal 8-bit image capture.
• Edge detection is achieved by comparing neighboring column data within in-column memories post-analog-to-digital conversion, eliminating the need for an external image signal processor.
• This on-chip integration facilitates high-speed and low-power edge image acquisition.

Key Insights:

• Successfully demonstrated edge image capture at a maximum frame rate of 60 fps.
• A prototype sensor with 1920 × 1440 resolution was fabricated using a 90-nm process.
• Achieved remarkably low power consumption of 9.4 mW at 60 fps.

Outlook:

• This technology holds promise for enhancing the efficiency of edge detection in low-power computer vision systems.
• Further development could lead to more sophisticated on-chip image processing capabilities.
• Potential applications include mobile devices, autonomous systems, and embedded vision platforms.