Deng Chenwei

Bio: Deng Chenwei is an academic researcher from Beijing Institute of Technology. The author has contributed to research in topics: Motion estimation & Second-generation wavelet transform. The author has an hindex of 2, co-authored 4 publications receiving 24 citations.

Recent advances and challenges of visual signal quality assessment

Abstract: While quality assessment is essential for testing, optimizing, benchmarking, monitoring, and inspecting related systems and services, it also plays an essential role in the design of virtually all visual signal processing and communication algorithms, as well as various related decision-making processes. In this paper, we first provide an overview of recently derived quality assessment approaches for traditional visual signals (i.e., 2D images/videos), with highlights for new trends (such as machine learning approaches). On the other hand, with the ongoing development of devices and multimedia services, newly emerged visual signals (e.g., mobile/3D videos) are becoming more and more popular. This work focuses on recent progresses of quality metrics, which have been reviewed for the newly emerged forms of visual signals, which include scalable and mobile videos, High Dynamic Range (HDR) images, image segmentation results, 3D images/videos, and retargeted images.

Image Quality Evaluation Method Based on Human Visual System

Abstract: MSE, PSNR, SSIM and other traditional objective image assessment metrics simply calculate sta- tistical average results between original images and recon- structed ones, which is incompatible with human visual system. To address this issue, a joint subjective and ob- jective image quality evaluation method is proposed. It has the following features: flrstly, the subjective feelings of re- constructed image are measured quantitatively in wavelet domain; secondly, only noticeable visual distortion is uti- lized in this model; lastly, based on difierent character- istics of spatio-temporal frequency response, eye-sensitive wavelet coe-cients are enhanced, while those non-sensitive ones are inhibited. Experimental results show that the pro- posed model has simple structure and is easy-to-operate; furthermore, it can re∞ect the visual efiects accurately and objectively.

Airport target detection algorithm in remote sensing images based on JPEG2000 compressed domain

Abstract: This paper mainly discusses airport target detection in remote sensing images. A new algorithm based on JPEG2000 compressed domain is proposed. Since the amount of remote sensing image data is generally large, images usually are compressed by JPEG2000. Therefore, this paper proposes an airport detection method for remote sensing image based JPEG2000 compressed domain. In the case of JPEG2000 compressed data with less decompression, we detect the airport target directly in the compressed domain. In the compressed domain, first extract the edge by canny edge operator, then detect parallel lines of airport runway and finally determine the target and fix the location of airport. By experiments, the algorithm proposed this paper can detect the airport target in JPEG2000 compressed domain. (4 pages)

Mesh-Based In-band Motion Estimation Method

Abstract: In this paper, we present a mesh-based motion estimation method in the wavelet domain. The Overcomplete Discrete Wavelet Transform (ODWT) is utilized to overcome the shift-variant property of the Discrete Wavelet Transform (DWT). Due to the good adaptation of non-translational motion, triangular mesh motion estimation is applied in the low frequency subband of the current frame. The initial motion vectors of other subbands are predicted from the low frequency subband. Experimental results show that our proposed method outperforms the other two methods in terms of Mean Absolute Difference (MAD) as well as the subjective quality.

Image Sharpness Assessment by Sparse Representation

Abstract: Recent advances in sparse representation show that overcomplete dictionaries learned from natural images can capture high-level features for image analysis. Since atoms in the dictionaries are typically edge patterns and image blur is characterized by the spread of edges, an overcomplete dictionary can be used to measure the extent of blur. Motivated by this, this paper presents a no-reference sparse representation-based image sharpness index. An overcomplete dictionary is first learned using natural images. The blurred image is then represented using the dictionary in a block manner, and block energy is computed using the sparse coefficients. The sharpness score is defined as the variance-normalized energy over a set of selected high-variance blocks, which is achieved by normalizing the total block energy using the sum of block variances. The proposed method is not sensitive to training images, so a universal dictionary can be used to evaluate the sharpness of images. Experiments on six public image quality databases demonstrate the advantages of the proposed method.

Full-Reference Image Quality Metrics: Classification and Evaluation (Foundations and Trends in Computer Graphics and Vision)

Abstract: The wide variety of distortions that images are subject to during acquisition, processing, storage, and reproduction can degrade their perceived quality. Since subjective evaluation is time-consuming, expensive, and resource-intensive, objective methods of evaluation have been proposed. One type of these methods, image quality (IQ) metrics, have become very popular and new metrics are proposed continuously. This paper aims to give a survey of one class of metrics, full-reference IQ metrics. First, these IQ metrics were classified into different groups. Second, further IQ metrics from each group were selected and evaluated against six state-of-the-art IQ databases.

Full-Reference Image Quality Metrics

Abstract: The wide variety of distortions that images are subject to during acquisition, processing, storage, and reproduction can degrade their perceived quality. Since subjective evaluation is time-consuming, expensive, and resource-intensive, objective methods of evaluation have been proposed. One type of these methods, image quality (IQ) metrics, have become very popular and new metrics are proposed continuously. This paper aims to give a survey of one class of metrics, full-reference IQ metrics. First, these IQ metrics were classified into different groups. Second, further IQ metrics from each group were selected and evaluated against six state-of-the-art IQ databases.

Recognizing Global Reservoirs From Landsat 8 Images: A Deep Learning Approach

Abstract: Man-made reservoirs are key components of terrestrial hydrological systems. Identifying the location and number of reservoirs is the premise for studying the impact of human activities on water resources and environmental changes. While complete bottom-up censuses can provide a comprehensive view of the reservoir landscape, they are time-consuming and laborious and are thus infeasible on a global scale. Moreover, it is challenging to distinguish man-made reservoirs from natural lakes in remote sensing images. This study proposes a convolutional neural network (CNN)-based framework to recognize global reservoirs from Landsat 8 imageries. On the basis of the HydroLAKES dataset, a Landsat 8 cloud-free mosaic of 2017 was clipped for each feature (reservoir or lake) and was resized into 224 × 224 patches, which were collected as training and testing samples. Compared to other deep learning methods (Alexnet and VGG) and state-of-the-art traditional machine learning methods (support vector machine, random forest, gradient boosting, and bag-of-visual-words), we found that fine-tuning the pretrained CNN model, ResNet-50, could reach the highest accuracy (91.45%). Application cases in Kansas (USA, North America), Mpumalanga (South Africa, Africa), and Kostanay (Kazakhstan, Asia) resulted in classification accuracies of better than 99%, which showed the applicability of the proposed ResNet-50 model to the extraction of reservoirs from a vast amount of moderate resolution images. The framework that was developed in this paper is the first attempt to combine remote sensing big data and the deep learning technique to the recognition of reservoirs at a global scale.

No- Reference Stereoscopic Image Quality Assessment

Abstract: We present two contributions in this work 1)a bi variate generalized Gaussian distribution (BGGD) model for the joint distribution of luminance and disparity sub band coefficients of natural Stereoscopic scenes. and 2) a no- reference (NR) stereo image quality assessment algorithm based on the BGGD model.