Image enhancement is a critical processing step for various vision-based applications, and many nonlinear methods are proposed. However, most of them are global techniques, lacking contrast enhancement to bring out fine features and details. Furthermore, they do not work well in various lighting conditions. In order to handle these limitations, an adaptive image enhancement method (AIELN) is proposed based on local luminance statistics and nonlinear functions. The method is composed of three steps: adaptive dynamic range adjustment, adaptive contrast enhancement and color restoration. Dynamic range adjustment is achieved by a series of nonlinear functions with different curvatures designed based on the human vision system, which can adaptively increase the intensity around dark regions and decrease the intensity around bright regions. Contrast enhancement is accomplished by enhancing the intensity of image according to the luminance of the local regions. Finally, the enhanced image is obtained by color restoration in YUV space. Experimental results demonstrate that the proposed method can be effectively improve the quality of the images captured in non-uniform lighting conditions, achieving better balance between dynamic range adjustment and contrast enhancement. Furthermore, the proposed method outperforms several existing methods in terms of quality and efficiency.
Just-noticeable difference (JND) is defined as the smallest intensity change in an image that can be noticed by the human vision system (HVS). Any perceptible distortion level must be greater than the JND. Based on this observation, a local binary pattern (LBP) is developed for image quality assessment. First, the JND map of the image is computed. The spatial and relative intensity relationships among pixels in a local neighborhood are employed to generate the proposed LBP based on the JND map. Then, image contrast is used as a weighting factor for the LBP histogram generation to characterize the structural and contrast information of the image. Finally, the contrast and structure changes due to image distortion are measured by calculating the similarity between contrast-weighted histograms of the reference and distorted images. Support vector regression is employed to pool the similarity to predict the quality. Experimental results on benchmark databases demonstrate that the proposed LBP can effectively and accurately measure image quality, which is consistent with the HVS. The proposed method achieves high consistency with subjective perception using 18 reference values and performs better than other state-of-the-art reduced reference image quality assessment methods.
High-resolution remote sounding instruments have thousands of channels usually, the similarity between channels will cause redundancy, and may lead to the non-convergence of retrieval algorithm. In this paper, the channel selection of temperature retrieval is discussed based on air-borne High-spectral resolution Interferometer Sounder. The results show that, for the total 1724 channels of HIS 600-1075cm-1 band, the selected 45 channels contain 70% of the total information content，97 channels contain 80% of the total information content, greatly reducing the number of channels involved in retrieval algorithm, then using selected channels simulated for the measurements of HIS is discussed, the average absolute deviation between the retrieved temperature profile and the truth is 0.74K/Km. The result demonstrates that stepwise iterative method is possible to select channels for the airborne high-spectral infrared data.
In space attack and defense, on-orbital servicing, pose estimation of unknown (non-cooperative) spacecrafts is one
of the most important conditions when taking the attack, defense and servicing measures. However, as for
non-cooperative spacecrafts, the imaging characteristics of the features and the geometric constraints among the features
are unknown, it is almost impossible to achieve target extraction, recognition, tracking, and pose solving automatically.
To solve this technical problem, the paper proposes a method to determine the pose of non-cooperative spacecrafts based
on collaboration of space-ground and rectangle feature. It employs a camera and rectangular features to achieve these
operations above mentioned automatically. Experimental results indicate that both the position errors and the attitude
errors satisfy the requirements of pose estimation during the tracking, approaching and flying round the non-cooperative
spacecraft. The method provides a new solution for pose estimation of the non-cooperative target, and has potential
significance for space-based attack and defense and on-orbital servicing.