Nanjing University of Information Science and Technology

About: Nanjing University of Information Science and Technology is a education organization based out in Nanjing, China. It is known for research contribution in the topics: Precipitation & Aerosol. The organization has 14129 authors who have published 17985 publications receiving 267578 citations. The organization is also known as: Nan Xin Da.

Efficient Minimum Error Bounded Particle Resampling L1 Tracker With Occlusion Detection

Abstract: Recently, sparse representation has been applied to visual tracking to find the target with the minimum reconstruction error from a target template subspace. Though effective, these L1 trackers require high computational costs due to numerous calculations for l1 minimization. In addition, the inherent occlusion insensitivity of the l1 minimization has not been fully characterized. In this paper, we propose an efficient L1 tracker, named bounded particle resampling (BPR)-L1 tracker, with a minimum error bound and occlusion detection. First, the minimum error bound is calculated from a linear least squares equation and serves as a guide for particle resampling in a particle filter (PF) framework. Most of the insignificant samples are removed before solving the computationally expensive l1 minimization in a two-step testing. The first step, named τ testing, compares the sample observation likelihood to an ordered set of thresholds to remove insignificant samples without loss of resampling precision. The second step, named max testing, identifies the largest sample probability relative to the target to further remove insignificant samples without altering the tracking result of the current frame. Though sacrificing minimal precision during resampling, max testing achieves significant speed up on top of τ testing. The BPR-L1 technique can also be beneficial to other trackers that have minimum error bounds in a PF framework, especially for trackers based on sparse representations. After the error-bound calculation, BPR-L1 performs occlusion detection by investigating the trivial coefficients in the l1 minimization. These coefficients, by design, contain rich information about image corruptions, including occlusion. Detected occlusions are then used to enhance the template updating. For evaluation, we conduct experiments on three video applications: biometrics (head movement, hand holding object, singers on stage), pedestrians (urban travel, hallway monitoring), and cars in traffic (wide area motion imagery, ground-mounted perspectives). The proposed BPR-L1 method demonstrates an excellent performance as compared with nine state-of-the-art trackers on eleven challenging benchmark sequences.

Modulation of the Tibetan Plateau Snow Cover on the ENSO Teleconnections: From the East Asian Summer Monsoon Perspective

Abstract: TheEastAsiansummermonsoon(EASM)mayexhibitratherlargevariabilitybetweenyearscharacterized by the same ENSO phase. Such inconsistency reduces the EASM predictability based on ENSO. Results in this study show that the Tibetan Plateau snow cover (TPSC) exerts a modulating effect on ENSO teleconnections and ENSO significantly correlates with the EASM only during the reduced TPSC summers. Three-dimensional circulation structures are examined to manifest that the typical ENSO signals in reduced TPSC summers tend to be stronger than in excessive TPSC summers. Numerical and theoretical evidences indicate that the anomalously reduced TPSC can force positive geopotential height anomalies at the upper troposphereandweakenthejetstreamsacrosseasternAsiaandnorthwesternPacific.Governedbysuchbasic state zonal flows, the extratropical Rossby wave response to the ENSO forcing usually has a larger amplitude and pronounced westward development. In such case, ENSO extends its influences to eastern Asia and enhances its connection with the EASM.

Fractal Dimensions and Mixing Structures of Soot Particles during Atmospheric Processing

Abstract: Soot particles strongly absorb sunlight and hence act as a short-lived warming agent. Atmospheric aging of soot particles changes their morphology and mixing state and consequently alters their optical properties. Here we collected soot particles at tunnel, urban, mountaintop, and background sites in the North China Plain and analyzed their mixing structures and morphology using transmission electron microscopy. Soot particles were further classified into three types: bare-like, partly coated, and embedded. Bare-like soot particles were dominant at the tunnel site, while most soot particles were of the partly coated or embedded type at other sites. Fractal dimensions (Df) of different types of soot particles ranged from 1.80 to 2.16 and increased in the following order: bare-like < partly coated < embedded. Moreover, their average Df changed from 1.8 to 2.0 from the tunnel to the background site. We conclude that the Df can characterize the shape of soot aggregates reasonably well, and its variation refle...

Coverless Image Steganography Without Embedding

Abstract: The traditional image steganography technologies designate a cover image and embed secret data into it to form the stego-image. However, the modification traces caused by the embedding will be left in the cover image, which makes successful steganalysis possible. To fundamentally address this issue, this paper proposes a novel image steganography framework, called as coverless image steganography, which does not need to employ the designated cover image for embedding the secret data. In this framework, a series of appropriate original images which already contain the secret data are directly chosen from a constructed database, and these original images can be regarded as the stego-images. Firstly, a number of images are collected to construct the database, and these images are indexed according to their hash sequence generated by a robust hashing algorithm. Then, the secret data is transformed to a bit string and divided into a number of segments. To implement the information hiding, the images of which the hash sequences are the same as the segments are chosen from the database as the stego-images. Experimental results show that the proposed coverless image steganography framework can resist the existing steganalysis tools, and have desirable robustness to the typical image attacks such as rescaling, luminance change, and noise adding.