Hyper Vision HPV-X2
High-Speed Video Camera
FTCMOS2 Advanced, Next-Generation Burst Image Sensor
Burst Method Enables Ultra High-Speed Recording
For typical high-speed video cameras, image storage memories are located outside of the image sensor. Because the number of signal output taps are overwhelmingly small compared to the number of pixels, the transfer of the video signals from the pixels to the memories must be a sequentially serial process; therefore, ultra high-speed recording of more than 1 million frames per second could not be realized. In contrast, Shimadzu's burst image sensor has the same number of built-in memories as number of frames recorded. Furthermore, a pixel and memories are connected by wire in a one-to-one manner in order to completely parallel transfer the video signal from the pixels to the memories. This makes it possible to realize ultra high-speed recording at 10 million frames per second. In addition, since it not limited to the number of signal output taps as with conventional serial transfer system, high-resolution recording at ultra high speed is available.
Next-Generation Burst Image Sensor Based on CMOS Technology
Conventional burst image sensors are based on CCD technology, in which the memory is positioned adjacent to the pixels. As a result, there are problems with decreased image quality due to signal leakage from pixels to memory. Accordingly, the Shimadzu FTCMOS burst image sensor adopts CMOS technology, in which the pixels and memory are spatially separated to achieve high image quality with no signal leaks.
In addition, with the FTCMOS2, light sensitivity is six times better than with FTCMOS, thanks to the adoption of a new CMOS process.
Note: FTCMOS and FTCMOS2 sensors were developed through collaborative research with Prof. Shigetoshi Sugawa of Tohoku University. Patents: 04931160, 04844853, 04844854