FPV Camera 1.2KM Ultra-Range recognition 1/2.6” 3.9mm 12mm Dual Cameras Multi-target locking Integrated Auto tracking&detection

The AI VisionCube Kit is a revolutionary AI visual tracking system designed specifically for drones, integrating high-performance target recognition, dynamic tracking, and anti-jamming control. Through the fusion of deep learning algorithms and high-performance intelligent chips, drones equipped with this kit can achieve automatic target acquisition, stable tracking, and precise trackings in complex electromagnetic environments. This significantly reduces the operational threshold for pilots and greatly improves mission execution efficiency.

Core functions and advantages

1.2km over-distance identification

It supports single/dual camera collaborative algorithm and high-precision optical modules. It can automatically identify human bodies, vehicles and other targets 1,200 meters away, which is more than 85% higher than the traditional AI image processing board’s recognition rate, which enables over-visual operation.

Multi-objective recognition

Based on YOLOv7 deep optimization, the multi-objective parallel processing architecture is synchronously identified up to 50 dynamic targets an

Easel in the picture

Using PiP dual-stream display technology, the global situational awareness of the main picture and the 8x electronic amplification detail tracking of the sub-picture are presented on the same screen, taking into account environmental analysis and accurate tracking decisions.

Adaptive adsorption tracking

After the drone recognizes multiple targets, heading correction can be made. The intelligent crosshairs adsorption algorithm will absorb the tracking target to the center of the screen, shortening the locking time to 0.5 seconds, greatly reducing the load of the pilot micro-operation, and increasing the locking success rate to 92% in high-speed mobile scenarios.

Dynamic memory lock

The fusion timing prediction model and the Kalman filtering algorithm are combined. After the target is temporarily blocked or removed from the field of view, it can still be recaptured within 3 seconds through motion trajectory memory, and the tracking stability in complex vegetation occlusion scenarios is increased by 40%.

Multi-spectral compatibility

Supports flexible switching between visible light/infrared dual-mode sensors, 9650 mV/lux·s ultra-high photosensitive performance and non-cooled infrared thermal imaging to achieve seamless combat capability on the day and night, and can accurately identify heat sources and camouflage targets in extreme environments -20℃ to 60℃.

Lightweight integration

Fully modular design, total weight <200g, compatible with open source flight controls such as Betaflight/APM, plug-and-play code-free debugging, complete drone installation within 5 minutes, adapted to industry-level six-rotor to small crossover aircraft, greatly reducing the cost of AI capability deployment.

Product Name: AI VisionCube S

Support detection target recognition type: default human and vehicle

Support target recognition distance: Car: 450m Person: 170m

Minimum tracking pixel of target object: 16*16 pixels

Whether picture-in-picture support:

AI image processing board main control computing power (TOPS): 1

AI Image Processing Board Control Protocol: CRSF

AI image processing board supports firmware: BetaFlight/APM

Input voltage (V): 9~16

AI image processing board size (length, width, height, mm): 38*38*27.7

AI image processing board installation hole position (mm): 25.5*25.5

Visible light camera sensor (Inch): 1/2.8

Visible camera focal length (mm): 4

Visible camera video input format: 1920*1080@30hz

——–

Accessories list:

AI VisionCube S x 1

Camera Base x 1

SH1.0-12PIN 15cm cable x 1 (for connecting flight control and AI VisionCube kit)

Cross slot disc head screw M2*4 x 2

Hexagon screw M2*6 x 4

Product Name: AI VisionCube D

Support target recognition distance: Car: 1200m Person: 500m

Minimum tracking pixel of target object: 16*16 pixels

Whether picture-in-picture support:

AI image processing board main control computing power (TOPS): 6

AI Image Processing Board Control Protocol: CRSF

AI image processing board supports firmware: BetaFlight/APM

AI image processing board input voltage (V): 9~16

AI image processing board size (length, width, height, mm): 38*38*26.7

AI image processing board installation hole position (mm): 25.5*25.5

Visible light camera 1 Sensor (Inch): 1/2.6

Visible camera 1 Focal length (mm): 3.9

Visible camera 1 Video input format: 1920*1080@60hz

Visible light camera 2 sensor L: 1/2.6 Inch

Visible camera 2 focal length (mm): 12

Visible camera 2 Video input format: 1920*1080@60hz

——–

Accessories list:

AI VisionCube D x 1

Camera Base x 1

SH1.0-12PIN 15cm cable x 1 (for connecting flight control and AI VisionCube kit)

Cross slot disc head screw M2*4 x 4

Hexagon screw M2*6 x 4

Product Name: AI VisionCube ST

Support target recognition distance: Car: 1200m Person: 500m

Minimum tracking pixel of target object: 16*16 pixels

Whether picture-in-picture support:

AI image processing board main control computing power (TOPS): 1

AI Image Processing Board Control Protocol: CRSF

AI image processing board supports firmware: BetaFlight/APM

AI image processing board input voltage (V): 9~16

AI image processing board size (length, width, height, mm): 38*38*27.7

AI image processing board installation hole position (mm): 25.5*25.5

Visible light camera 1 Sensor (Inch): 1/2.8

Visible camera 1 Focal length (mm): 4

Visible camera 1 Video input format: 1920*1080@30hz

Thermal imaging camera Focal length: 9.1mm

Thermal imaging camera Response band: 8~14μm (LWIR)

Thermal imaging camera Video input format: 384*288@25hz

Thermal imaging camera Interface: USB

——–

Accessories list:

AI VisionCube ST x 1

Camera Base x 1

SH1.0-12PIN 15cm cable x 1 (for connecting flight control and AI VisionCube kit)

Cross slot disc head screw M2*4 x 2

Hexagon screw M2*6 x 4

Product Name: AI VisionCube DT

Support target recognition distance: Car: 1200m Person: 500m

Minimum tracking pixel of target object: 16*16 pixels

Whether picture-in-picture support:

AI image processing board main control computing power (TOPS): 6

AI Image Processing Board Control Protocol: CRSF

AI image processing board supports firmware: BetaFlight/APM

AI image processing board input voltage (V): 9~16

AI image processing board size (length, width, height, mm): 38*38*26.7

AI image processing board installation hole position (mm): 25.5*25.5

Visible light camera 1 Sensor (Inch): 1/2.6

Visible camera 1 Focal length (mm): 3.9

Visible camera 1 Video input format: 1920*1080@60hz

Visible light camera 2 sensor L: 1/2.6 Inch

Visible camera 2 focal length (mm): 12

Visible camera 2 Video input format: 1920*1080@60hz

Visible camera 2 Camera size: 40.8*25*26

Thermal imaging camera Focal length: 9.1mm

Thermal imaging camera Response band: 8~14μm (LWIR)

Thermal imaging camera Video input format: 384*288@25hz

Thermal imaging camera Interface: USB

——

Accessories list:

AI VisionCube DT x 1

Camera Base x 1

SH1.0-12PIN 15cm cable x 1 (for connecting flight control and AI VisionCube kit)

Cross slot disc head screw M2*4 x 4

Hexagon screw M2*6 x 4