Author: |
Researchers at Changchun Institute of Optics, Fine Mechanics and Physics (CIOMP) has successfully developed a prototype of the hyperspectral camera, it was revealed at the recently closed Light Conference 2019.
The prototype, based on pixel-level splitting technology, was completed in April, according to CIOMP Associate Researcher Dr. HU Haihong, who gave a speech on "Pixel-level tunable hyperspectral video intelligent imaging technology" at Light Conference 2019.
Camera prototype(Photo by CIOMP)
According to Dr. HU Haihong, the project leader, the prototype is a staring hyperspectral camera that is light weight, small in size, has a simple structure that is highly stable, and needs no extra data calculation. This camera can achieve high frame rate staring imaging and getting hyperspectral imaging which have spectrum band data in single exposure. In addition, the camera have high spatial resolution, high time resolution, and high signal to noise ration hyperspectral video imaging.
Pixel spectrogram under the microscope (Photo by CIOMP)
The prototype was used to successfully identify camouflage clothes and plants, performing as well as European and US products from leading manufacturers in the field.
Hyperspectral video cameras based on pixel-level spectroscopic technology were only available from a few overseas companies such as IMEC and Silios in Europe, and Pixelteq in the United States. CIOMP’s prototype is the first domestic product in China.
Traditionally, China’s hyperspectral imaging technology was based on prism/grating spectroscopic, interference or linear gradient filter and other spectroscopic systems, which use push-broom or whisk-broom mode, which made hyperspectral video-level imaging impossible.
The HAN Chengshan project team and WANG Xiaoyi project team of CIMOP, who developed the prototype camera, have overcome many difficulties and solved several technical problems, such as high-spectral pixel-level spectroscopic, pixel-level optics, integration, high-frequency high-speed electronics design and spectral data inversion recovery to achieve their success.
Imaging recognition experiment (Photo by CIOMP)
中文
