Due to the enormous volume of point cloud data, transmitting and storing data requires large bandwidth storage space. It could be a critical bottleneck, especially in tasks such as autonomous driving. In this letter, we propose novel compression algorithm based on clustering. The proposed scheme starts with range image-based segmentation step, which segments three-dimensional (3-D) into ground main objects. Then, it introduces prediction method according segmented regions' shape. This is...

LiDAR sensors are almost indispensable for autonomous robots to perceive the surrounding environment. However, transmission of large-scale point clouds is highly bandwidth-intensive, which can easily lead problems, especially unstable communication networks. Meanwhile, existing data compression mainly based on rate-distortion optimization, ignores semantic information ordered and task requirements robots. To address these challenges, this article presents a task-driven Scene-Aware Point...

High quality image reconstruction from undersampled <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">k</i> -space data is key to accelerating MR scanning. Current deep learning methods are limited by the small receptive fields in networks, which restrict exploitation of long-range information, and impede mitigation full-image artifacts, particularly 3D tasks. Additionally, substantial computational demands considerably hinder advancements...

In this letter, we propose a novel coding architecture for LiDAR point cloud sequences based on clustering and prediction neural networks. clouds are structured, which provides an opportunity to convert the 3D data 2D array, represented as range images. Thus, cast compression images problem. Inspired by high efficiency video (HEVC) algorithm, design sequence. The scans divided into two categories: intra-frames inter-frames. For intra-frames, cluster-based intra-prediction technique is...

Due to the huge volume of point cloud data, storing or transmitting it is currently difficult and expensive in autonomous driving. Learning from high efficiency video coding (HEVC) framework, we propose an advanced scheme for large-scale LiDAR sequences, which several techniques have been developed remove spatial temporal redundancy. The proposed strategy consists mainly intra-coding inter-coding. For intra-coding, utilize a cluster-based prediction method inter-coding, predictive recurrent...

Light detection and ranging (LiDAR) plays an indispensable role in autonomous driving technologies, such as localization, map building, navigation object avoidance. However, due to the vast amount of data, transmission storage could become important bottleneck. In this article, we propose a novel compression architecture for multi-line LiDAR point cloud sequences based on clustering convolutional long short-term memory (LSTM) networks. clouds are structured, which provides opportunity...

With the development of video technology, a large amount data generated from conferences, sports events, live broadcasts and network classes flows into our daily lives. However, ultra-high-definition transmission is still challenge due to limited bandwidth instability, which further affects quality service closely linked with consumer electronic display. To address this challenge, we propose deep-learned perceptual control approach, can significantly improve visual experience at same...

With complementary multi-modal information (i.e. visible and thermal), multispectral pedestrian detection is essential for around-the-clock applications, such as autonomous driving, video surveillance, vicinagearth security. Despite its broad the requirements expensive thermal device multi-sensor alignment limit utilization in real-world applications. In this paper, we propose a pseudo-multispectral (called PseudoMPD) method, which employs gray image converted from RGB to replace real image,...

As the latest video coding standard, high‐efficiency (HEVC) achieves better performance and supports higher resolution compared with predecessor H.264/advanced (AVC). Intra‐coding is an important feature in HEVC which reduces spatial redundancy significantly, due to flexible structure, high density of angular prediction modes. However, improvement on efficiency obtained at expense extraordinary computation complexity. This study presents a novel unit (CU) partitioning technique for HEVC. By...

In this paper, we propose a novel perceptual-based intra coding optimization algorithm for the High Efficiency Video Coding (HEVC) using deep convolution networks (DCNs). According to saliency map, can intelligently adjust bit rate allocation between salient and non-salient regions of video. The proposed strategy mainly consists two techniques, map extraction, intelligent allocation. First, train DCN model generate that highlights semantically regions. Compared with texture-based region...

Due to the huge volume of point cloud data, storing and transmitting it is currently difficult expensive in autonomous driving. Learning from high-efficiency video coding (HEVC) framework, we propose a novel compression scheme for large-scale sequences, which several techniques have been developed remove spatial temporal redundancy. The proposed strategy consists mainly three parts: intracoding, intercoding, residual data coding. For inspired by depth modeling modes (DMMs), 3-D HEVC...

In this paper a fusion model by combining the Stationary Wavelet Transform (SWT), Quantum Genetic Algorithm (QGA) and Back-propagation (BP) Neural Network is proposed to forecast wireless network traffic. order achieve guaranteed Quality of Service (QoS) in networks, various managing measures can be taken only knowing traffic advance. This developed which called SWT-QGA-BP efficiently used assess future provide adequate evidence for management. By using SWT, original non-stationary data are...

In this article, we present an efficient coding scheme for LiDAR point cloud maps. As a map consists of numerous single scans spliced together, by recording the time stamp and quaternion matrix each scan during building, cast compression into sequence problem. The architecture includes two techniques: intra-coding inter-coding. For intra-frames, segmentation-based intra-prediction technique is developed. inter-frames, interpolation-based inter-frame network explored to remove temporal...

Cognitive radio technology has experienced a fast development since its basic concept was raised by Joseph Mitola. And wireless cognitive sensor network is new application field of Radio based on the traditional network. The senor node assigned ability spectrum sensing, data analysis and parameter adjusting can function as node. On one side, sense environment, detect primary users take use vacant spectrum. other functions arouse some energy security concerns. Malicious which reports false...