Real Infrared Image Sequences Research Articles

Improved neural network based scene-adaptive nonuniformity correction method for infrared focal plane arrays

An improved scene-adaptive nonuniformity correction (NUC) algorithm for infrared focal plane arrays (IRFPAs) is proposed. This method simultaneously estimates the infrared detectors' parameters and eliminates the nonuniformity causing fixed pattern noise (FPN) by using a neural network (NN) approach. In the learning process of neuron parameter estimation, the traditional LMS algorithm is substituted with the newly presented variable step size (VSS) normalized least-mean square (NLMS) based adaptive filtering algorithm, which yields faster convergence, smaller misadjustment, and lower computational cost. In addition, a new NN structure is designed to estimate the desired target value, which promotes the calibration precision considerably. The proposed NUC method reaches high correction performance, which is validated by the experimental results quantitatively tested with a simulative testing sequence and a real infrared image sequence.

Robust super-resolution by minimizing a Gaussian-weighted L2 error norm

Super-resolution restoration is the problem of restoring a high-resolution scene from multiple degraded low-resolution images under motion. Due to imaging blur and noise, this problem is ill-posed. Additional constraints such as smoothness of the solution via regularization is often required to obtain a stable solution. While adding a regularization term to the cost function is a standard practice in image restoration, we propose a restoration algorithm that does not require this extra regularization term. The robustness of the algorithm is achieved by a Gaussian-weighted L2-norm in the data misfit term that does not response to intensity outliers. With the outliers suppressed, our solution behaves similarly to a maximum-likelihood solution in the presence of Gaussian noise. The effectiveness of our algorithm is demonstrated with super-resolution restoration of real infrared image sequences under severe aliasing and intensity outliers.

Improved Kalman-filter nonuniformity correction algorithm for infrared focal plane arrays

An improved nonuniformity correction algorithm based on the Kalman-filter is presented. This paper gives an introduction to a piecewise linear model of the detector response curve, and the nonuniformity correction algorithm based on a Kalman-filter is extended and improved, which is suitable for infrared focal plane arrays with nonlinearity of the response characteristic. Compared to the original algorithm, it can resolve not only the problems of detector offset and gain drift with time, but also the influence of the detector nonlinear response to the nonuniformity correction performance to a certain extent, so it provides better nonuniformity correction. The NUC ability of the improved algorithm is demonstrated with real infrared image sequences.

Moving dim point target detection with three-dimensional wide-to-exact search directional filtering

We assess the performance of a novel three-dimensional wide-to-exact search directional filtering (3DWESDF) algorithm for detecting and tracking a weak moving dim target against a cluttered background in real infrared (IR) image sequence. This paper proposed a novel 3DWESDF to decrease the 3D directional filter’s (3DDF) computational requirements and increase the target energy accumulation ability further. Prior to the filtering, a newly pre-whitening method termed three-dimensional spatial-temporal adaptive prediction filter (3DSTAPF) used here to suppress complex cluttered background. Extensive experimental results demonstrate the proposed algorithm’s ability in detecting weak dim point target against a complex cloud-cluttered background in real IR image sequence. Finally, performance comparisons of the proposed algorithm and 3DDF, on real IR image data, are presented in which the advantages of the proposed 3DWESDF filters are shown.

Novel dim target detection and estimation algorithm based on double threshold partial differential equation

The aim of the present work is to propose a brand-new algorithm based on an adaptive double threshold nonlinear anisotropic diffusion equation (DTPDE) to detect and track moving dim targets against complex cluttered background in infrared (IR) image sequences. We also illustrate the performance comparisons of the proposed algorithm DTPDE and two-dimensional least mean squares (TDLMS) on real IR image sequence data. Extensive experiment results demonstrate the proposed novel algorithm's flexibility and adaptability in detecting moving weak dim targets.

Point target detection in IR image sequences: a hypothesis-testing approach based on target and clutter temporal profile modeling

We approach the problem of point target detection in infrared image sequences by modeling the temporal behavior of clutter and targets on a single-pixel basis. These models, which are experimentally verified, are then used to develop a temporal likelihood-ratio test and derive the corresponding decision rule. We demonstrate the effectiveness of the technique by applying it to real infrared image sequences containing targets of opportunity and evolving cloud clutter. The physical models and resulting hypothesis-testing approach could also be applicable to other image-sequence-processing scenarios, using acquisition systems besides infrared imaging, such as the detection of small moving objects or structures in a biomedical or biological imaging scenario, or the detection of satellites, meteors, or other celestial bodies in night-sky imagery acquired using a telescope.