We as humans are very good at understanding an image; fast and accurate. For computers, this is rather challenging. Various techniques using hand-crafted pixel-based image descriptors are u. The last couple of years has shown drastic changes in image classification area. Deep convolutional neural networks (CNNs) have been effectively used in many image-related tasks.
CNNs were first introduced in the 90s. Despite some promising results, a wider adoption was not observed. This was mostly due to compute intensive requirements and not extending to other visual tasks. With recent improvements in graphical processing units, it’s now possible to create bigger networks and use big data within realistic time.
At ArgosAI we focus on image understanding. Our main focus is recognizing foreign object debris (FOD) in any weather condition on the airport runway surface. For this purpose we created our own image recognition solution using deep neural networks. Since we deal with high stake decisions human level accuracy is the primary concern. We achieve that using our state of the art technology and long hours of computational operations.
Get in touch for a detailed discussion and see how we can help you make your flights safer.
#FOD #neuralnetworks #deeplearning #flightsafety