High-resolution ISAR imaging based on photonic receiving for high-accuracy automatic target recognition.

Area of Science:

• Radar Systems Engineering
• Photonics and Optical Engineering
• Signal Processing

Background:

• Traditional Inverse Synthetic Aperture Radar (ISAR) imaging faces limitations in resolution and detection range.
• Photonic sampling offers a novel approach to overcome bandwidth and fidelity constraints in radar signal acquisition.

Purpose of the Study:

• To demonstrate a high-resolution ISAR imaging scheme utilizing photonic receiving technology.
• To evaluate the performance of the developed ISAR system in terms of image resolution and automatic target recognition (ATR) capabilities.

Main Methods:

• Implementation of an ISAR imaging scheme employing a photonic analog-to-digital converter (PADC) for direct sampling of linear frequency modulated (LFM) pulse echoes.
• Acquisition of ISAR images for complex targets, specifically two pyramids, to assess 2D resolution.
• Application of deep learning algorithms for ATR on the generated high-resolution ISAR dataset.

Main Results:

• Achieved a two-dimensional (2D) resolution of 3.3 cm × 1.9 cm for complex targets.
• Demonstrated ISAR imaging without detection range swath limitations due to high-fidelity photonic sampling.
• Validated a 95% accuracy for ATR of three complex targets using a small deep learning training dataset.

Conclusions:

• Photonic receiving technology enables high-resolution ISAR imaging with significant improvements in resolution.
• The developed ISAR system shows promise for effective ATR applications, even with limited training data.
• This approach overcomes traditional limitations, paving the way for advanced radar imaging and target identification.