14 Civil Marine Radar modes with ambiguous parametric parameters from a Synthetic Radar Dataset Generator. Expressed as images for image classification in AI methods. Authors: Richard Rudd-Orthner and Lyudmila Mihaylova. The 14 radar mode and model combinations are: S-band Kelvin Hughes SharpEye 24NM Fast Scan S-band Kelvin Hughes SharpEye 24NM Slow Scan S-band Kelvin Hughes SharpEye 48NM Fast Scan S-band Kelvin Hughes SharpEye 48NM Slow Scan S-band Kelvin Hughes SharpEye 96NM Fast Scan S-band Kelvin Hughes SharpEye 96NM Slow Scan X-band Kelvin Hughes SharpEye 96NM Fast Scan X-band Kelvin Hughes SharpEye 96NM Slow Scan X-band Kelvin Hughes 1262 SharpEye 24NM Fast Scan X-band Kelvin Hughes 1262 SharpEye 24NM Slow Scan X-band Kelvin Hughes 1262 SharpEye 48NM Fast Scan X-band Kelvin Hughes 1262 SharpEye 48NM Slow Scan RayMarine HD RayMarine Quantum For human review: The Images are ".png" format of 3 colours 368 wide by 251 Height pixels each, with a bicubic filter applied for the human interpretation, and the ".avi" files are the movies from those images. For the dataset in the machine, and without the bicubic filter: X (input image) and Y (output classification) data are in ".npy" TensorFlow 2.0 format as in np.load (filename). Xdata = an image of 100 by 499 pixels in 3 colours which is 10ms lines over 1 second at 50Khz sampling. The Red channel is a spectrogram and the blue and green are the I/Q phase multiplied by the magnitude. Ydata = The radar mode and model tag Accuracy in image classification is 99.8%