Kernel is a modular camera platform that can be easily connected together to form custom, compact camera arrays. Each camera module is a self-contained, open source Linux computer with a single sensor and on-board storage. Modules connect together with link boards or breakout cables which distribute power and signals, allowing a single port for array communications.
We spent years during the Kernel R&D phase modifying other company's cameras and listening to our customers. The biggest complaint was that end users wanted the ability to modify and repair their camera but the manufacturers had not designed the product to be easily adjusted. Kernel is designed to be easily taken apart and its components configured to the user's project requirements. You can easily swap sensors, lenses and filters in each camera module and configure the camera software so the image's embedded EXIF metadata matches the hardware configuration. If a new sensor comes out you can simply purchase just the sensor board and not the entire new camera. This reduces overall product costs while allowing your existing product to stay updated. If a certain component in the camera becomes damaged and needs to be replaced outside of the warranty, the end user can purchase just that part and easily make the repair.
Since Kernel is capable of capturing images quickly we also designed it to be the most reliable in continuous shutter control. Kernel supports PWM trigger and for additional accuracy you can use a relay (voltage) pulse. Connecting to an autopilot's UAVCAN port allows Kernel to receive the 50Hz EKF data stream containing the GNSS location. Features such as UNIX and GNSS nanosecond time-stamps, number of leap seconds and UAVCAN precision and fix status are also saved in each image's metadata. For exposure moment detection Kernel also sends out a relay pulse, so if you currently use a hot-shoe detection PPK system Kernel will work the same. The camera also calculates the time difference between the trigger and exposure moments and writes that delay time to each captured image.
Kernel is ideal for vehicle (e.g., aerial drone) manufacturers, integrators, and operators looking to build powerful, custom, multi-spectral, and varying field of view camera arrays.
|Processor:||Freescale i.MX 6 Dual Core ARM Cortex A9 1.2GHz|
|Operating System:||Yocto Linux (Open Source)|
|Sensor Options:||1.2 MP mono/RGB (global), 3.2 MP mono/RGB (global), 14.4 MP RGB (rolling)|
|Image Format:||12bit RAW, 16bit TIFF (per channel)|
|Frames Per Second (FPS):||9fps (1.2MP RAW), 7fps (3.2MP RAW), 2fps (14.4MP RAW)|
|Lenses:||43mm 47° HFOV (1.2MP sensor), 47mm 41° HFOV (3.2MP & 14.4MP sensors), ANY LENS (CS/C mount, all sensors)|
|Ground Sample Distance (GSD) Resolution (120m/400ft AGL):||7.5cm/px (1.2MP), 4.3cm/px (3.2MP), 2.0cm/px (14.4MP)|
|Shutter Trigger:||PWM, Relay pulse (high-low)|
|Exposure Feedback:||Relay pulse (high-low)|
|Filter Options (Mono Sensors):||405nm, 440nm, 450nm, 490nm, 518nm, 550nm, 590nm, 615nm, 632nm, 650nm, 725nm, 780nm, 808nm, 850nm, 880nm|
|Filter Options (RGB Sensors):||Visible Light (IR Cut), 395nm+870nm, 475nm+850nm, 550nm+850nm, 660nm+850nm, 475nm+550nm+850nm, 550nm+660nm+850nm|
|Connectivity:||USB2.0, UART, UAVCAN, PWM (in and out), I2C, Ethernet, GPIO, HDMI & SD Video|
|Positional Sensor:||Inertial Measurement Unit (IMU) - 3xAccel., 3xAng., 3xMag. (ST Micro LSM9DS1TR )|
|Expansion Ports:||Side 40-pin “Kernel Port”, Bottom 60-pin "Expansion Port"|
|Memory Storage:||Removable microSDXC (up to 128 GB)|
|Power (Kernel):||5.0 VDC, 4.0 W (max)|
|Dimensions (Kernel):||34.0 mm x 34.0 mm x 40+ mm (length x width x height)|
|Weight (Kernel):||45g+ (depends on sensor+lens chosen)|