Full color night vision Camera AI ISP Tech for DJI M300 drone

The full-color night-vision camera for DJI M300 drone, it is based on the DJI SDK payload. With the AI ISP technology, the gimbal camera can achieve full-color night vision effect under extremely low light conditions, breaking through the black and white vision, no fear of darkness, ultra-clear imaging, no reflection but no exposure, and super fog remove.

Features:

• Based on DJI SDK Payload, compatible with DJI Matrice 300 RTK drone.
• 20X optical zoom,1/1.8 “super star level CMOS, 0.0001Lux.
• Ultra-long-distance target detection, dynamic strong light suppression.
• No need for fill light, no infrared laser, completely rely on camera and AI ISP technology to realize full-color view in the dark.
• Based on the analysis of very weak light (visible light: 380-790 nm) environmental image signals, a set of deep learning that can separate noise and real signals from very weak light signals is trained by using neural networks to deeply learn the distribution characteristics of noise and signals. The AI algorithm enhances the real signal to the normal light environment intensity while reducing noise, and the signal-to-noise ratio can be improved by up to 25dB, realizing normal imaging in extremely low light environment.
• On the basis of extremely low-light imaging technology, it analyzes and uses neural network to deeply learn the color characteristics of all-weather image signals, and can obtain high-definition, high-quality full-color images even in extremely low-light environments (illumination reaches 0.001lux).
• The super wide dynamic technology uses neural network to deeply learn the regional division of multiple exposure image signals, and fuses the image data of long, medium and short frames with different exposure times, and integrates the dark and bright areas of the image. Without losing, the color details of the target object are captured in extremely low illumination. It can reach up to 64 times the original dynamic range, and the dynamic range can be improved by up to 36dB, which is 300% higher than the industry. The backlight scene is clear, and the dark light scene is full of color, so that the visual perception will never be afraid of backlight.
• The night vision autofocus technology uses AI to identify the characteristics of the target in different illumination scenes, and adaptively adjusts the focusing algorithm and parameters according to the target, achieving unassisted and fast focusing under ultra-low illumination, which is completely impossible in existing night vision products. It can also achieve normal focusing in dark scenes.
• Dehazing technology uses deep learning algorithm and artificial neural network to learn the characteristics of fog in real scenes, and identifies and removes fog of different degrees in real time, so as to achieve color and fog-free picture output, and achieve high contrast and high definition of the overall picture.
• In night vision scenarios, traditional cameras will obtain better imaging results by increasing the exposure time or synthesizing multiple frames. Using AI technology such as extremely low-light imaging, night vision effect is achieved without increasing exposure, which solves the smear phenomenon of traditional night vision

Specs: