Uses wide-angle, ultra-wide-angle, telephoto, macro, or depth sensors to acquire front, upper, lower, left, and right view information.
A multi-faceted prism or multiple mirrors divide and guide rear upper, lower, left, and right rays.
The acquired images are integrated into a complete view through matching, boundary correction, 3D reconstruction, and reprojection.
This flow shows rear rays being acquired and integrated into a complete view while switching between prism and mirror methods.
Combines front and side information with divided rear images and outputs the result as a cubemap or panorama.
A multi-faceted prism is placed at the geometric center of four lenses to separately acquire the rear upper, lower, left, and right areas. Because it does not block the front optical axis, spatial efficiency is high.
Upper, lower, left, and right mirrors with different angles are arranged to guide rear rays. There is no chromatic aberration, and reflection angles can be independently designed for each area.
The second optical group is spaced so that it is not positioned on the front optical axis.
The rear upper, lower, left, and right views are separately acquired and integrated.
Time synchronization is secured through a global-shutter simultaneous exposure structure.
Performance can be improved through AP, NPU, GPU, or app updates.
Applied as an auxiliary optical module that secures a full view including front and rear views.
Supports robots in recognizing objects behind and to the side as well as in front.
Reduces blind spots of vehicles, drones, and moving bodies and supports multidirectional environmental recognition.
Used as a surveillance structure that checks a wide range and rear area simultaneously from a single device.
A high-precision mirror-based optical structure can be applied to equipment requiring rear or side observation.
Expanded into inspection devices for areas difficult to view directly, such as the back side, inside, or narrow spaces of products.