RfHelicopterDetector neural net v1.0 – neural network for detecting helicopters in video frames. The network is two files: RfHelicopterDetector.cfg (neural network configuration file) и RfHelicopterDetector.weights (neural network weights file). The neural network detects helicopters on video frames and generates a list of detected objects (helicopters). Information about each object includes its position on the video frame (for each detected object a bounding box is calculated), as well as the probability that the image of the object belongs to the "helicopter" class.


SCOPE OF APPLICATION

A neural network for helicopter detection is used in tactical detection and destruction systems, in flight safety systems to prevent aircraft collisions, in helicopter landing automation systems, and also in perimeter protection systems of special importance ground objects.


TARGET PLATFORMS AND COMPATIBILITY

For video processing using a neural network, the following frameworks are used: PyTorch, Keras, Tensorflow and Darknet. The compatibility of the neural network with a specific operating system and hardware platform is determined by the framework used. Neural network files can be provided on request to evaluate performance on a specific computing platform.


ADAPTATION OF NEURAL NETWORK

In case the quality of the network requires improvement for specific application conditions we can modify the neural network for best objects detection, if you have enough video material to conduct research. We can also modify the network for use on low-power computing platforms.


PRICE AND TERMS OF DELIVERY

Neural network files are delivered under the perpetual SCBL license (license of SOLID Ltd. for compiled versions of software) without the right to transfer to third parties. The price of a perpetual license for neural network is 6000 EUR.


LOADS

  • RfHelicopterDetector neural net v1.0 Brochure [DOWNLOAD]

  • Software source code license (SSCL) [DOWNLOAD]


QUALITY ASSESSMENT

You have the opportunity to assess the neural network performance on your video. Due to the fact that the RfHelicopterDetector neural net files cannot be safely included into the executable file of the free demo program distributed, customer’s video processing to assess the network performance is performed exclusively by Solid ltd. on request.


write to us

Name *
Name