This page was machine-translated and may differ from the original. View original
Small pixels supporting VGA+ (640x600) and HD+ (1124x1364)
Implementing unrivaled BSI global shutter pixels with die stacking technology
STMicroelectronics has launched a new high-speed image sensor to support next-generation smart computer vision applications.
Use Global Shutter mode, which is primarily used for distortion-free image capture when moving scenes or near-infrared lighting are required.
Rolling shutters, which capture pixel data one line at a time, often result in distorted moving images or require additional correction work, whereas global shutter sensors can store all pixel data of each frame simultaneously.
▲ Global-shutter machine-vision sensors <Image=ST>
The sensors introduced this time are the VD55G0 with 640 x 640 pixels and the VD56G3 with 1.5M pixels (1124 x 1364), achieving the smallest resolutions on the market at 2.6mm x 2.5mm and 3.6mm x 4.3mm.
Low inter-pixel crosstalk is possible across all wavelengths, including near-infrared, ensuring clear images with high contrast. The optical flow processing function built into the VD56G3 can calculate motion vectors without processing on a host computer, making it suitable for various applications such as augmented and virtual reality (AR/VR), SLAM (Simultaneous Localization and Mapping), and 3D scanning.
ST's advanced pixel technologies, including full Deep Trench Isolation (DTI), enable ultra-small 2.61㎛ x 2.61㎛ pixels on a single wafer layer, combining low Parasitic Light Sensitivity (PLS), excellent Quantum Efficiency (QE), and low crosstalk.
In addition, by implementing small pixels on a back-illuminated (BSI) wafer and vertically stacking optical sensors and related signal processing circuits at the bottom of the die, space can be reduced while minimizing sensor size and embedding key functions such as a fully autonomous optical flow block.
The bottom die is manufactured using ST's 400nm technology and integrates digital and analog circuits. The high-density, low-power digital circuits are equipped with all hardware features, including an exposure algorithm for statistics collected from up to 336 regions, as well as automatic defect correction and automatic dark correction.
The fully autonomous low-power optical flow block can calculate 2,000 motion vectors at a speed of 60fps. The built-in vector generation function is very useful for augmented and virtual reality or robotics, and supports SLAM or 6DoF (Six Degrees of Freedom) applications, especially on host systems with limited processing power.
This sensor supports multi-frame context through sequencing programming functions as well as full lighting control, and is equipped with an integrated temperature sensor, I2C high-speed mode+ control, fault correction, windowing, binning, and MIPI CSI-2 data interface.
Implementing unrivaled BSI global shutter pixels with die stacking technology
STMicroelectronics has launched a new high-speed image sensor to support next-generation smart computer vision applications.
Use Global Shutter mode, which is primarily used for distortion-free image capture when moving scenes or near-infrared lighting are required.
Rolling shutters, which capture pixel data one line at a time, often result in distorted moving images or require additional correction work, whereas global shutter sensors can store all pixel data of each frame simultaneously.

▲ Global-shutter machine-vision sensors <Image=ST>
The sensors introduced this time are the VD55G0 with 640 x 640 pixels and the VD56G3 with 1.5M pixels (1124 x 1364), achieving the smallest resolutions on the market at 2.6mm x 2.5mm and 3.6mm x 4.3mm.
Low inter-pixel crosstalk is possible across all wavelengths, including near-infrared, ensuring clear images with high contrast. The optical flow processing function built into the VD56G3 can calculate motion vectors without processing on a host computer, making it suitable for various applications such as augmented and virtual reality (AR/VR), SLAM (Simultaneous Localization and Mapping), and 3D scanning.
ST's advanced pixel technologies, including full Deep Trench Isolation (DTI), enable ultra-small 2.61㎛ x 2.61㎛ pixels on a single wafer layer, combining low Parasitic Light Sensitivity (PLS), excellent Quantum Efficiency (QE), and low crosstalk.
In addition, by implementing small pixels on a back-illuminated (BSI) wafer and vertically stacking optical sensors and related signal processing circuits at the bottom of the die, space can be reduced while minimizing sensor size and embedding key functions such as a fully autonomous optical flow block.
The bottom die is manufactured using ST's 400nm technology and integrates digital and analog circuits. The high-density, low-power digital circuits are equipped with all hardware features, including an exposure algorithm for statistics collected from up to 336 regions, as well as automatic defect correction and automatic dark correction.
The fully autonomous low-power optical flow block can calculate 2,000 motion vectors at a speed of 60fps. The built-in vector generation function is very useful for augmented and virtual reality or robotics, and supports SLAM or 6DoF (Six Degrees of Freedom) applications, especially on host systems with limited processing power.
This sensor supports multi-frame context through sequencing programming functions as well as full lighting control, and is equipped with an integrated temperature sensor, I2C high-speed mode+ control, fault correction, windowing, binning, and MIPI CSI-2 data interface.
본 기사에 대한 정정·반론·추후보도 청구는 보도 청구 안내를, 그간 게재된 보도문은 정정·반론보도 모아보기를 참고해 주세요.














