Neuromorphic Cameras in Astronomy: Unveiling the Future of Celestial Imaging Beyond Conventional Limits

TL;DR

Neuromorphic cameras address the limitations of conventional frame-based optical astronomy by using asynchronous event-driven sensing and a logarithmic response, achieving HDR $>100\,\mathrm{dB}$ and timing at the scale of $\mu$s. Through deployments on a $1300$ mm DFOT and a $200$ mm Dobsonian, the work demonstrates photometry of faint stars, HDR imaging of the Trapezium under full Moon, and high-temporal-resolution imaging of satellites and meteoroids, with flux calibration via GAIA transformations. These results provide a practical pathway to integrate neuromorphic sensing into astronomy, enabling rapid transient studies, improved bright-source handling, and potentially reducing data volumes for next-generation surveys. Future work includes calibration standardization, algorithm development for event-to-photometry conversion, and exploration of synergies with adaptive optics and traditional detectors.

Abstract

To deepen our understanding of optical astronomy, we must advance imaging technology to overcome conventional frame-based cameras' limited dynamic range and temporal resolution. Our Perspective paper examines how neuromorphic cameras can effectively address these challenges. Drawing inspiration from the human retina, neuromorphic cameras excel in speed and high dynamic range by utilizing asynchronous pixel operation and logarithmic photocurrent conversion, making them highly effective for celestial imaging. We use 1300 mm terrestrial telescope to demonstrate the neuromorphic camera's ability to simultaneously capture faint and bright celestial sources while preventing saturation effects. We illustrate its photometric capabilities through aperture photometry of a star field with faint stars. Detection of the faint gas cloud structure of the Trapezium cluster during a full moon night highlights the camera's high dynamic range, effectively mitigating static glare from lunar illumination. Our investigations also include detecting meteorite passing near the Moon and Earth, as well as imaging satellites and anthropogenic debris with exceptionally high temporal resolution using a 200mm telescope. Our observations show the immense potential of neuromorphic cameras in advancing astronomical optical imaging and pushing the boundaries of observational astronomy.

Figures (9)

• Figure 1: Working principles of neuromorphic and conventional cameras. (a) Structure of a neuromorphic camera pixel, with components analogous to the human retina: a photodiode (photoreceptor), difference amplifier (bipolar cells), and comparators (on-off ganglion cells), including the transformation of light intensity into the logarithmic domain and the spike generation process. (b) Asynchronous data transmission in the neuromorphic camera using AER readout. (c) The high temporal resolution of the neuromorphic camera enables tracking of the fast-rotating dot without motion blur. (d) Conventional camera pixel components: photodiode, charge storage, shutter switch, and gain unit. (e) Conventional camera workflow showing digital conversion by the ADC and frame generation. (f) Example of a fast-rotating dot in a static background. Frame generation at 30fps leads to motion blur. (g) Illustration of a full moon night sky showcasing the neuromorphic camera's ability to capture faint objects like rocket debris and dim stars with minimal interference from bright moonlight and accurately tracking the fast-moving space station. In contrast, the conventional camera's output is significantly impacted by intense moonlight and motion blur, obscuring faint objects and blurring the space station.
• Figure 2: Simulation-based comparison of point source imaging using a conventional and a neuromorphic camera: (a) In a conventional camera, light is integrated over the exposure duration, producing a Gaussian-shaped intensity distribution. The intensity profiles along the horizontal and vertical lines passing through the centre of the star exhibit a Gaussian nature. (b) A neuromorphic camera, in contrast, generates events based on brightness changes caused by atmospheric tip/tilt variations. This results in a doughnut-shaped output after event integration, with distinct intensity distributions along the horizontal and vertical centre lines compared to the conventional camera.
• Figure 3: Observational setup for neuromorphic camera-based celestial observations using the 1300 mm DFOT telescope at Devasthal, ARIES, India. (a) The neuromorphic camera installed at the focal plane of the telescope. (b) Roll-off rooftop moving back to expose the telescope to the night sky. (c) Close-up view of the neuromorphic camera mounted on telescope using a custom interface plate. (d) The neuromorphic camera integrated with the interface plate.
• Figure 4: Celestial Observations with the neuromorphic camera on the 1300 mm DFOT. False colors are used to emphasize variations in pixel intensity within the grayscale images for visualization purposes. Images captured include: (a) HIP 9884, V-band magnitude 2, 10s acquisition. (b) Zoomed view of HIP 9884 showing a donut-shaped profile and the corresponding intensity profile along a marked red line. (c) Jupiter and its moon, Ganymede, 10s V-band acquisition. (d) Zoomed view of Ganymede with its intensity profile along a marked red line. (e) The multiple star system SAO 97646, 6″ separation, 10s V-band acquisition. (f) Zoomed view of SAO 97646, displaying the intensity profile along the red line. (g) SAO 92721, V-band magnitude 5, 20s acquisition. (h) Zoomed view of a star in SAO 92721 with the corresponding intensity profile along the red line.
• Figure 5: Signal processing steps followed to perform photometric analysis on event data captured from the neuromorphic camera.