Abstract:The EHT has captured a series of images of black holes. These images could provide valuable information about the gravitational environment near the event horizon. However, accurate detection and parameter estimation for candidate black holes are necessary. This paper explores the potential for identifying black holes in the ultraviolet band using space telescopes. We establish a data pipeline for generating simulated observations and present an ensemble neural network model for black hole detection and parameter estimation. The model achieves mean average precision [0.5] values of 0.9176 even when reaching the imaging FWHM ($\theta_c$) and maintains the detection ability until $0.54\theta_c$. The parameter estimation is also accurate. These results indicate that our methodology enables super-resolution recognition. Moreover, the model successfully detects the shadow of M87* from background noise and other celestial bodies and estimates its inclination and positional angle. Our work demonstrates the feasibility of detecting black holes in the ultraviolet band and provides a new method for black hole detection and further parameter estimation.
Instrumentation and Methods for Astrophysics,General Relativity and Quantum Cosmology
What problem does this paper attempt to address?
### Problems the paper attempts to solve
The paper aims to solve the problem of identifying black holes in the ultraviolet band through space telescopes, and proposes a deep - learning - based method to detect and estimate black - hole parameters. Specifically, the paper explores the potential of using space telescopes to identify black holes in the ultraviolet band, and establishes a data - generation pipeline and an integrated neural - network model for black - hole detection and parameter estimation.
### Main research contents
1. **Data - generation pipeline**:
- **Black - hole accretion - disk simulation**: Use the ray - tracing method under general relativity to simulate the accretion disk around the black hole and generate black - hole shadow images.
- **Telescope simulation**: According to the configuration of the Hubble Space Telescope, simulate the point - spread function (PSF) of the telescope, and convolve the PSF with the simulated image to obtain the observation results.
- **Noise simulation**: Consider the signal - to - noise ratio (SNR) of the telescope, add Gaussian noise to make the simulated observation results closer to the actual observations.
2. **Integrated neural - network model**:
- **Detection model**: Use the YOLO model for black - hole detection, which can accurately distinguish black holes from observation images containing dozens to hundreds of other celestial bodies and determine their positions.
- **Parameter - estimation model**: Use the EfficientNet - b1 model for black - hole parameter estimation, including the inclination angle \( i \), mass \( M \), position angle \( \phi \) and accretion - disk temperature \( T \).
### Key formulas
1. **Event horizon of Schwarzschild black hole**:
\[
r_s=\frac{2GM}{c^2}
\]
where \( G \) is the gravitational constant, \( c \) is the speed of light, and \( M \) is the mass of the black hole.
2. **Accretion - disk temperature**:
\[
T = 6.27\times10^7\left(\frac{\alpha}{0.1}\right)^{1/4}\left(\frac{M}{3M_\odot}\right)^{-3/4}\left(\frac{\dot{M}}{10^{17}\text{g/s}}\right)^{1/2}\text{K}
\]
where \( M_\odot \) is the solar mass, \( \dot{M} \) is the accretion rate, and \( \alpha \) is the standard α - viscosity coefficient.
3. **Full - width - at - half - maximum (FWHM) of telescope imaging**:
\[
\theta_c=\frac{1.22\lambda}{D}
\]
where \( \lambda \) is the wavelength and \( D \) is the diameter of the telescope.
4. **Signal - to - noise ratio (SNR)**:
\[
\text{SNR}=\frac{S_oQt}{\sqrt{S_oQt + S_bQtn_p+S_dtn_p+R^2n_p}}
\]
where \( S_o \) is the number of photoelectrons from the target, \( S_b \) is the number of photoelectrons from the sky background, \( S_d \) is the number of photoelectrons generated by the dark current, \( R \) is the read - out noise, \( t \) is the exposure time, \( Q \) is the quantum efficiency of the CCD, and \( n_p \) is the number of pixels occupied by the target.
### Experimental verification
The paper tests the effectiveness of the model by using the image of M87* and the observation data of the Hubble Space Telescope. The results show that the average precision of the model in the detection task reaches 0.9176, and it can maintain the detection ability even within the imaging FWHM range. For the parameter - estimation task, the model can accurately recover the inclination angle, position angle, accretion - disk temperature and black - hole mass.
### Conclusion
The paper demonstrates the feasibility of detecting black holes in the ultraviolet band and provides a new method to detect black holes and further estimate their parameters. By combining deep - learning and optical - simulation techniques, this method can exceed the traditional Rayleigh diffraction limit and achieve...