In this paper, a target and background driven simulation procedure is developed for optimal band analysis and performance evaluation of multispectral sensors for dim target detection, in which the observation geometry, target and background radiant characteristics, and the influence of infrared sensor system are integrated into. With the specified sensor altitude, the BeiHang University - Atmospheric Transfer Model (BHU-ATM) is adopted to calculate spectral irradiances of space-variant sky background. When a target with an assumed altitude exists in one line of sight (LOS), the corresponding nadir angle is used to calculate the distance between the target and the sensor, which impacts the target spectral irradiance at the sensor aperture. To analyze the optimal band for target detection, a set of spectral response functions with different central wavelengths and bandwidths are designed to calculate the target-to-background contrasts as well as the signal-to-noise ratios (SNRs). To demonstrate the usefulness of the developed procedure, typical sensor parameters are used to analyze the optimal band for aircraft detection. The band achieving the highest SNR is selected and used in the radiant image simulation for performance evaluation. The results show that the detection performance is related to spectral band as well as the LOS direction.
The intensive emission of earth limb in the field of view of sensors contributes much to the observation images. Due to the low signal-to-noise ratio (SNR), it is a challenge to detect small targets in earth limb background, especially for the detection of point-like targets from a single frame. To improve the target detection, track before detection (TBD) based on the frame sequence is performed. In this paper, a new technique is proposed to determine the target associated trajectories, which jointly carries out background removing, maximum value projection (MVP) and Hough transform. The background of the bright earth limb in the observation images is removed according to the profile characteristics. For a moving target, the corresponding pixels in the MVP image are shifting approximately regularly in time sequence. And the target trajectory is determined by Hough transform according to the pixel characteristics of the target and the clutter and noise. Comparing with traditional frame-by-frame methods, determining associated trajectories from MVP reduces the computation load. Numerical simulations are presented to demonstrate the effectiveness of the approach proposed.
In this work, a fast calculation method of the scattered radiance for scenario involving both the earth surface and the earth-limb regions is proposed. The single scattering equation of typical two-stream approximate is adapted to compute atmospheric radiative transfer under the spherical-parallel atmosphere assumption. With specified atmospheric profiles, spectral band and observation geometry, a two-dimensional (2-D) matrix of the scattered radiance varying with incident zenith and viewing zenith angles are then calculated. Finally, the earth disk images are generated for different spectral bands by interpolating the calculated radiance matrices. Simulation results of multispectral earth disk images for space-based earth observation sensors are presented to demonstrate the usefulness of the proposed technique for high fidelity scene generation where both the earth surface and the earth-limb regions are observed.
In this work, we focus on developing the infrared (IR) sensor and performance analysis model for space-based IR
systems which are designed for detection of space targets in the earth and the earth-limb background. Corresponding to
the sensor observation geometry, a simplified transmittance calculation scheme applicable to large-scale scenes as well
as a mathematical model for pixel-by-pixel irradiance calculation is proposed. By defining the apparent contrast of
targets in simulated IR images, a model for detection performance analysis is developed for sensors operating in different
spectral bands. Typical simulation examples are presented to validate the current model and methodology.