Chongliang Liu

Robust multiframe super-resolution reconstruction based on regularization

Images of high-resolution are desired and often required in most photoelectronic imaging applications, and corresponding image restoration algorithm has became the frontier research. A novel multiframe super-resolution reconstruction algorithm based on stochastic regularization is proposed in this paper. By analyzing the image degradation model, the iterative gradient method based on Taylor series expansion is applied in the algorithm to estimate the inter-frame displacement. The L1 norm is used for fusing the data of low-resolution frames and removing outliers, and the regularization technique based on bilateral total variation is used to remove artifacts from the final answer and improve the rate of convergence. Simulated and real experiment results confirm the effectiveness of the algorithm.

An IR Image Gray-Scale Transformation Algorithm Based on Human Visual System

For human observation and advanced processing in actual applications, the gray-scale range expansion and scenery contrast improvement are the key problems of IR image processing. Aiming at the problems, by analyzing the IR image characters and human visual system, an IR image gray-scale transformation algorithm based on human visual system (HVSBGT) is proposed. The algorithm is combined threshold division and non-linear gray-scale transformation. The experiments of actual IR images from staring cooled and un-cooled medium wave IRFPA thermal imagers show that the proposed algorithm is effective on expanding the gray-scale range and improving the scenery contrast. Using the algorithm, the IR image entropy is increased and better than those using traditional algorithms. Meanwhile, by setting different thresholds, the sky background is inhibited and some object details are enhanced partially such as buildings, water surface and plants. For digital signal processor (DSP) hardware processing, a large number of exponential and logarithmic operations are simplified by using gray-scale mapping table. The HVSBGT algorithm can provide high quality IR images for subjective observation and lay a foundation for advanced algorithms. The realization of HVSBGT algorithm is of great value and practical significance for the thermal imaging systems and equipments.

A self-adaptive nonuniformity correction algorithm for infrared images combined with two-point correction along the rim

Nonuniformity correction plays an important role in improving image quality of the infrared focal plane array. A self-adaptive nonuniformity correction algorithm for infrared images combined with two-point correction along the rim is proposed in this paper. With the help of perimeter blackbody in low and high temperature, two-point correction is executed initially to perimeter detectors. Then based on the scene information and shift between adjacent frames, a special algebraic algorithm is proposed to transport correction parameters from perimeter detectors to those interior un-corrected ones. In this way, the correction parameters of the whole field of view (FOV) are calculated. All calculations are algebraic, with a low computation load. This algorithm can realize adaptive two-point correction without covering the central FOV rapidly. It proves to get a satisfying correction effect after being tested by using real infrared images.

A DSP-based neural network non-uniformity correction algorithm for IRFPA

An effective neural network non-uniformity correction (NUC) algorithm based on DSP is proposed in this paper. The non-uniform response in infrared focal plane array (IRFPA) detectors produces corrupted images with a fixed-pattern noise(FPN).We introduced and analyzed the artificial neural network scene-based non-uniformity correction (SBNUC) algorithm. A design of DSP-based NUC development platform for IRFPA is described. The DSP hardware platform designed is of low power consumption, with 32-bit fixed point DSP TMS320DM643 as the kernel processor. The dependability and expansibility of the software have been improved by DSP/BIOS real-time operating system and Reference Framework 5. In order to realize real-time performance, the calibration parameters update is set at a lower task priority then video input and output in DSP/BIOS. In this way, calibration parameters updating will not affect video streams. The work flow of the system and the strategy of real-time realization are introduced. Experiments on real infrared imaging sequences demonstrate that this algorithm requires only a few frames to obtain high quality corrections. It is computationally efficient and suitable for all kinds of non-uniformity.