Low Probability of Intercept (LPI) radar waveform has complex and diverse modulation schemes, which cannot be easily identified by the traditional methods. The research on intrapulse modulation LPI radar waveform recognition has received increasing attention. In this paper, we propose an automatic LPI radar waveform recognition algorithm that uses a multi-resolution fusion convolutional neural network. First, signals embedded within the noise are processed using Choi-William Distribution (CWD) to obtain time-frequency feature images. Then, the images are resized by interpolation and sent to the proposed network for training and identification. The network takes a dual-channel CNN structure to obtain features at different resolutions and makes features fusion by using the concatenation and Inception module. Extensive simulations are carried out on twelve types of LPI radar waveforms, including BPSK, Costas, Frank, LFM, P1~P4, and T1~T4, corrupted with additive white Gaussian noise of SNR from 10dB to -8dB. The results show that the overall recognition rate of the proposed algorithm reaches 95.1% when the SNR is -6dB. We also try various sample selection methods related to the recognition task of the system. The conclusion is that reducing the samples with SNR above 2dB or below -8dB can effectively improve the training speed of the network while maintaining recognition accuracy.

This paper reports a high-quality hardware-friendly integer motion estimation (IME) scheme. According to different characteristics of CTU content, the proposed method adopts different adaptive multi-resolution strategies coupled with accurate full-PU modes IME at the finest level. Besides, by using motion vector derivation, IME for the second reference frame is simplified and hardware resource is saved greatly through processing element (PE) sharing. It is shown that the proposed architecture can support the real-time processing of 4K-UHD @60fps, while the BD-rate is just increased by 0.53%.

This research develops a method for trajectory planning of robotic systems with differential constraints based on hierarchical partitioning of a continuous state space. Unlike conventional roadmaps which is constructed in the configuration space, the proposed state roadmap also includes additional state information, such as velocity and orientation. A bounded domain of the additional state is partitioned into sub-intervals with multiple resolution levels. Each node of a state roadmap consists of a fixed position and an interval of additional state values. A valid transition is defined between a pair of nodes if any combination of additional states, within their respective intervals, produces a trajectory that satisfies a set of safety constraints. In this manner, a trajectory connecting arbitrary start and goal states subject to safety constraints can be obtained by applying a graph search technique on the state roadmap. The hierarchical nature of the state roadmap reduces the computational cost of roadmap construction, the required storage size of computed roadmaps, as well as the computational cost of path planning. The state roadmap method is evaluated in the trajectory planning examples of an omni-directional mobile robot and a car-like robot with collision avoidance and various types of constraints.

The conventional bilateral motion estimation (BME) for motion-compensated frame rate up-conversion (MC-FRUC) can avoid the problem of overlapped areas and holes but usually results in lots of inaccurate motion vectors (MVs) since 1) the MV of an object between the previous and following frames is more likely to have no temporal symmetry with respect to the target block of the interpolated frame and 2) the repetitive patterns existing in video frame lead to the problem of mismatch due to the lack of the interpolated block. In this paper, a new BME algorithm with a low computational complexity is proposed to resolve the above problems. The proposed algorithm incorporates multi-resolution search into BME, since it can easily utilize the MV consistency between two adjacent pyramid levels and spatial neighboring MVs to correct the inaccurate MVs resulting from no temporal symmetry while guaranteeing low computational cost. Besides, the multi-resolution search uses the fast wavelet transform to construct the wavelet pyramid, which not only can guarantee low computational complexity but also can reserve the high-frequency components of image at each level while sub-sampling. The high-frequency components are used to regularize the traditional block matching criterion for reducing the probability of mismatch in BME. Experiments show that the proposed algorithm can significantly improve both the objective and subjective quality of the interpolated frame with low computational complexity, and provide the better performance than the existing BME algorithms.

Traditional face swapping technologies require that the faces of source images and target images have similar pose and appearance (usually frontal). For overcoming this limit in applications this paper presents a pose-free face swapping method based on personalized 3D face modeling. By using a deformable 3D shape morphable model, a photo-realistic 3D face is reconstructed from a single frontal view image. With the aid of the generated 3D face, a virtual source image of the person with the same pose as the target face can be rendered, which is used as a source image for face swapping. To solve the problem of illumination difference between the target face and the source face, a color transfer merging method is proposed. It outperforms the original color transfer method in dealing with the illumination gap problem. An experiment shows that the proposed face reconstruction method is fast and efficient. In addition, we have conducted experiments of face swapping in a variety of scenarios such as children's story book, role play, and face de-identification stripping facial information used for identification, and promising results have been obtained.

The computing of applications in embedded devices suffers tight constraints on computation and energy resources. Thus, it is important that applications running on these resource-constrained devices are aware of the energy constraint and are able to execute efficiently. The existing execution time and energy profiling tools could help developers to identify the bottlenecks of applications. However, the profiling tools need large space to store detailed profiling data at runtime, which is a hard demand upon embedded devices. In this article, a reconfigurable multi-resolution profiling (RMP) approach is proposed to handle this issue on embedded devices. It first instruments all profiling points into source code of the target application and framework. Developers can narrow down the causes of bottleneck by adjusting the profiling scope using the configuration tool step by step without recompiling the profiled targets. RMP has been implemented as an open source tool on Android systems. Experiment results show that the required log space using RMP for a web browser application is 25 times smaller than that of Android debug class, and the profiling error rate of execution time is proven 24 times lower than that of debug class. Besides, the CPU and memory overheads of RMP are only 5% and 6.53% for the browsing scenario, respectively.

Concomitantly with the progress of wireless communications, cognitive radio has attracted attention as a solution for depleted frequency bands. Cognitive radio is suitable for wireless sensor networks because it reduces collisions and thereby achieves energy-efficient communication. To make cognitive radio practical, we propose a low-power multi-resolution spectrum sensing (MRSS) architecture that has flexibility in sensing frequency bands. The conventional MRSS scheme consumes much power and can be adapted only slightly to process scaling because it comprises analog circuits. In contrast, the proposed architecture carries out signal processing in a digital domain and can detect occupied frequency bands at multiple resolutions and with low power. Our digital MRSS module can be implemented in 180-nm and 65-nm CMOS processes using Verilog-HDL. We confirmed that the processes respectively dissipate 9.97 mW and 3.45 mW.

This paper proposes a novel image inpainting method based on bandelet transform. This technique is based on a multi-resolution layer to perform image restoration, and mainly utilizes the geometrical flow of the neighboring texture of the damaged regions as the basis of restoration. By performing the warp transform with geometrical flows, it transforms the textural variation into the nearing domain axis utilizing the bandelet decomposition method to decompose the non-relative textures into different bands, and then combines them with the affine search method to perform image restoration. The experimental results show that the proposed method can simplify the complexity of the repair decision method and improve the quality of HVS, and thus, repaired results to contain the image of contour of high change, and in addition, offer a texture image of high-frequency variation. These repair results can lead to state-of-the-art results.

The 3GPP Long Term Evolution Advanced (LTE-A) system, as compared to the LTE system, is anticipated to include several new features and enhancements, such as the usage of channel bandwidth beyond 20 MHz (up 100 MHz), higher order multiple input multiple output (MIMO) for both downlink and uplink transmissions, larger capacity especially for cell edge user equipment, and voice over IP (VoIP) users, and wider coverage and etc. This paper presents some key enabling technologies including flexible uplink access schemes, advanced uplink MIMO receiver designs, cell search, adaptive hybrid ARQ, and multi-resolution MIMO precoding, for the LTE-A system.

We present the use of a Modified Generalized Hough Transform (MGHT) and deformable contours for image data retrieval where a given contour, gray-scale, or color template image can be detected in the target image, irrespective of its position, size, rotation, and smooth deformation transformations. Potential template positions are found in the target image using our novel modified Generalized Hough Transform method that takes measurements from the template features by extending a line from each edge contour point in its gradient direction to the other end of the object. The gradient difference is used to create a relationship with the orientation and length of this line segment. Potential matching positions in the target image are then searched by also extending a line from each target edge point to another end along the normal, then looking up the measurements data from the template image. Positions with high votes become candidate positions. Each candidate position is used to find a match by allowing the template to undergo a contour transformation. The deformed template contour is matched with the target by measuring the similarity in contour tangent direction and the smoothness of the matching vector. The deformation parameters are then updated via a Bayesian algorithm to find the best match. To avoid getting stuck in a local minimum solution, a novel coarse-and-fine model for contour matching is included. Results are presented for real images of several kinds including bin picking and fingerprint identification.

Digital watermarking techniques were developed for regular raster data such as images or video, but little research addressed irregular vector data, such as the shapes of cartoons or elevation contours. Vector graphic images, such as those in SVG format, are popular on the WWW, and provide the advantage of permitting affine transformations without aliasing. The creation of cartoon images or the acquisition of GIS geometry data involves much work, so the copyright and ownership of vector data must be protected. Common components in vector graphic images are polygonal lines or polylines. This work develops a normal multi-resolution representation of a polygonal line, and embeds a copyright notice or serial number in this representation. Previous studies on polyline watermarking have the non-transparent problems, including self-intersection of line segments. The experimental results demonstrate that the proposed watermarking approach is perceptually transparent, and solves the self-intersection problem. It is also resistant to similarity transformation, traversal reordering, point insertion/deletion and random noise attacks.

Adaptive array antennas, which control their own patterns by means of feed-back or feed-forward control, are effective tools for gain enhancement and interference suppression. However, when applying them to mobile terminals, the problems of hardware complexity and power consumption need to be taken into consideration. One solution is the use of analog device-based adaptive array antennas, such as Reactively Steered Adaptive Array (RESAA) antennas and phased array antennas, which have the attractive characteristics of low cost and power consumption. In this paper, we propose an adaptive beamforming method based on a one-dimension search algorithm for phased array antennas with Micro Electro Mechanical Systems (MEMS) phase shifters, taking into consideration their slow operating speed due to mechanical structure of the devices. Furthermore, a smoothing processing is introduced to prevent the effect of noise and a multi-resolution alogrithm is proposed to help the system form beams more quickly and stably. Numerical results based on the IEEE 802.11a Wireless Local Area Network (WLAN) standard show that the proposed method has good interference suppression and gain enhancement capabilities in multipath fading channels.

In this paper, we propose a fast multi-resolution block matching algorithm with three resolution levels (upper, middle, and lower levels) for multiple-frame motion estimation (MFME). The main concept of the algorithm is to perform a fast search while maintaining a PSNR performance similar to a full search block matching algorithm (FSBMA). The algorithm combines motion vector prediction using the spatial correlation of motion vectors and a multiple candidate search based on a multi-resolution search. To further reduce the computational complexity, we propose two temporal reduction schemes. To reduce the number of previous reference frames to be processed, the first scheme is applied to the upper level by using the information obtained from the search results of the spatio-temporally adjacent macroblocks (MBs) and the result from the current MB in the middle level of the first reference frame. The other scheme is applied to the lower level by using statistical information. Experimental results show that the proposed algorithm guarantees an average PSNR loss of less than 0.23 dB with dramatically reduced computational complexity as compared to the FSBMA. In particular, for sequences with fast motion or frame skipping, the proposed method provides a more prominent PSNR performance than those of existing fast schemes with a comparable computational complexity.

Currently, the automatic image inpainting methods emphasize the inpainting techniques either globally or locally. They didn't consider the merits of global and local techniques to compensate each other. On the contrary, the artists fixed an image in global view firstly, and then focus on the local features of it, when they repaired it. This paper proposes a progressive processing of image inpainting method based on multi-resolution analysis. In damaged and defective area, we imitate the artistic techniques to approach the effectiveness of image inpainting in human vision. First, we use the multi-resolution characteristics of wavelet transform, from the lowest spatial-frequency layer to the higher one, to analyze the image from global-area to local-area progressively. Then, we utilize the variance of the energy of wavelet coefficients within each image block, to decide the priority of inpainting blocks. Finally, we extract the multi-resolution features of each block. We take account of the correlation among horizontal, vertical and diagonal directions, to determine the inpainting strategy for filling image pixels and approximate a high-quality image inpainting to human vision. In our experiments, the performance of the proposed method is superior to the existing methods.

This paper presents a multi-resolution optical flow estimation method that is robust against large variation in the estimation parameter. For each level solution of the multi-grid estimation, a nonlinear iteration is proposed differently from the existing method, where the incremental displacement from the coarser level optical flow is calculated by linear iteration. The experimental results show that the proposed scheme has better error-performance in a much wider range of regularization parameters.

This paper proposed a watermarking algorithm for image, which assumed an image compression based on DWT (Discrete Wavelet Transform). To reduce the amount of computation, this algorithm selects the watermarking positions by a threshold table which is statistically established from computing the energy correlation of the corresponding wavelet coefficients. The proposed algorithm can operate in a real-time if the image compression process operates in a real-time because the watermarking process was designed to operate in parallel with the compression process. Also it improves the property of losing the watermak and reducing the compresson ratio by the quantization and Huffman coding steps. It was done by considering the sign of the coefficients and the change in the value for watermarking. Visually recognizable pattern such as a binary image were used as the watermark. The experimental results showed that the proposed algorithm satisfied the properties of robustness and imperceptibility that are the major conditions of watermarking.

Conventional vector quantization (VQ) encoding method by full search (FS) is very heavy computationally but it can reach the best PSNR. In order to speed up the encoding process, many fast search methods have been developed. Base on the concept of multi-resolutions, the FS equivalent fast search methods using mean-type pyramid data structure have been proposed already in. In this Letter, an enhanced sum pyramid data structure is suggested to improve search efficiency further, which benefits from (1) exact computing in integer form, (2) one more 2-dimensional new resolution and (3) an optimal pair selecting way for constructing the new resolution. Experimental results show that a lot of codewords can be rejected efficiently by using this added new resolution that features lower dimensions and earlier difference check order.

Adapting to the structure of 2-D H-Transform, this paper proposes a novel wavelet domain half-pixel motion compensation algorithm HMRME (Half-pixel Multi-Resolution Motion Estimation). The primary objective of this study is the reduction of the aliasing effect caused by the down-sampling in the wavelet transform under the complexity constraints. The conventional multi-resolution motion estimation scheme can be combined with the half-pixel interpolation method to generate a new high-performance wavelet video codec. The preliminary results show that the performance of HMRME rises above its counterparts, the Multi-Resolution Motion Estimation (MRME) and the Adaptive Multi-Resolution Motion Estimation (AMRME).

The constant-Q based wavelet transform is the most effective means of quantitatively characterizing high frequency transient signals. This study develops a novel non constant-Q based multi-resolution transform (NCQM) and provides a precision analysis descriptor for both low and high frequency transients. The properties of this novel NCQM kernel are thoroughly examined and then the striking conceptual resemblance, energy conservation characteristic, and power spectrum close forms are derived. The rapid algorithm of NCQM is also presented and its excellent performance in noisy environments is demonstrated.

Recently a high-resolution image that has more than one million pixels is available easily. However, such an image requires much processing time and memory for an image understanding system. In this paper, we propose an integrated image understanding system of multi-resolution analysis and multi-agent-based architecture for high-resolution images. The system we propose in this paper has capability to treat with a high-resolution image effectively without much extra cost. We implemented an experimental system for images of indoor scenes.