Meihua Xu

A Novel Method of Pedestrian Detection Aided by Color Self-similarity Feature

Pedestrian detection has been widely applied in intelligent surveillance and driver assistant systems. The histogram of the oriented gradient (HOG) is the most commonly used feature in pedestrian detection algorithms, which is computationally intensive and results in slow detection speed. This paper proposes a method of pedestrian detection, which is based on color self-similarity (CSS) feature and AdaBoost classifier. The color self-similarity (CSS) feature calculates the ratio of two rectangles to measure the self-similarity in HSV color space, and then the AdaBoost classifier is used to screen out the detection windows containing pedestrian. Tests show that this method has the same detection accuracy and faster detection speed compared with HOG detectors.

A Novel Real-Time Pedestrian Detection System on Monocular Vision

Accuracy and speed are the two important keys in pedestrian detection. In order to balance these two indexes well, this thesis presents a novel pedestrian detection system, ROIs cascaded Uniform LBP and improved HOG, for real-time pedestrian detection in monocular vision. Two contributions are made in this system. First contribution is that Uniform LBP (Local Binary Pattern) cascaded improved HOG (Histograms of Oriented Gradients) are the novel structure for pedestrian detection, which can improve detection speed. Second contribution is that this pedestrian detection system is evaluated by many methods and algorithms. Experiment shows that this system can deal with 31 fps, which can be used as the real time pedestrian detection system.

A novel medical internet of things perception system based on visual image encryption and intrusion detection

Smart cities are the emerging topics in the recent decade with advancements in science and technology especially in the areas of communication technologies. Concept of smart cities depicts the inteconnectivity of physical objects to provide transparency in communication between the physical components or objects. However, implementation of a smart city using Internet of Things (IoT) which is the backbone architecture is affected by several attacks in the wireless communication domain during communication of information between the components as well as cyber-attacks which necessiate the need for the intrusion detection mechanism to prevent any illegeal loss data. This research paper addresses these security issues by implementing the IoT architecture and systematically analysing the demands in security aspects as well as justifying the results with experimentations conducted in a simulated smart city environment.

A Two-Stage Optimal Detection Algorithm Research for Pedestrians in Front of the Vehicles

In this paper, a two-stage optimal detection algorithm is presented for pedestrians in front of the vehicles. It uses the idea of combing the coarse-grain and fine-grain to effectively classify and filter. First, it uses the combination of Color Self-Similarity features based on rectangular block summing and AdaBoost classifier based on greedy strategy to coarse-grained screen the pedestrian detection window, then it uses the combination of HOG feature and libsvm classifier to fine-grained confirm the previous screened pedestrian detection window, Finally, the target windows is integrated by the greedy strategy. The AdaBoost classifier’s training time is theoretically shorten to the 1/T time of original algorithm. With the training process, The Color Self-Similarity features shorten to 250 dimensions by the feature selection. Then, the method makes full use of the image information and ensures the detection accuracy.

Improvement of Non-maximum Suppression in Pedestrian Detection Based on HOG Features

Pedestrian detection is a hot topic in the field of computer vision in recent year. But the current studies about pedestrian detection mainly focus on feature extraction, training and classifier model and pay little attention to non-maximum suppression (NMS). This thesis uses the information like ratio of detection scores, neighborhood window to improve NMS based on HOG-SVM algorithm, solving the problems that alone windows in detected images arise false detection rate and the suppression windows surrounded by inhibited windows arise false detection rate and missing detection rate. Experiment results on the INRIA pedestrian database show that the improved non-maxima suppression can solve the above problems, reducing the false detection rate and missing detection rate in pedestrian detection.

CORDIC-based parameters-fusion HOG IP for extracting feature

Histograms of Oriented Gradients (HOG) descriptor significantly outperforms other features for object detection, especially for pedestrian detection. Exploitation of high performance and efficient HOG IP has been a research hotspot in Automotive Electronic. In this letter, a CORDIC-based Parameters-fusion LUT HOG algorithm is proposed and made as the hardware IP (Intelligent Property). To decrease the computational effort and improve the accuracy, pipeline Coordinate Rotation Digital Computer (CORDIC) is applied to generate the magnitude and orientation of gradient. Parametersfusion Look-Up-Table (LUT) based regional division can operate the tri-linear interpolation in HOG with high speed. Circuit design and chip fabrication were performed using 0.18 m  CMOS technology. Measurement result shows that this CORDIC-based Parameters-fusion LUT HOG IP reduces the hardware overloads and can be used as the general IP for extracting features.

Analysis of Information Encryption on Electric Communication Network

The greatly improvement of electric power automation is making electric power system increasingly depend on the information networks to ensure its safety, reliable and efficient operation. This paper introduces the general situation of electric power system and power information network system in nowadays. Finally, two typical encryption algorithms DES and RSA are analyzed and compared, and software program based on QT is used to prove the principle of the encryption algorithms. This paper also has some reference value for the research on security of electric power telecommunication.

QVGA OLED Display Control Module with High Gray-Level

OLED has a lot of advantages like simple structure, ultrathin, self-luminescence, high brightness, short response time, wide viewing angle, low operation voltage and so on, which is widely applied in cell phone, PDA, DC, on board display and military field. This paper represents a real-time video display system on OLED based on the detailed analysis of OLED panel electrical characteristics and various gray scale scanning principles of the OLED scan and drive circuit. FPGA is the core control device in the whole system, the DVI decoded signal is processed and real-time displayed on the OLED. 240×RGB (H)×320 (V) QVGA, 256 gray scale is implemented and frame frequency is 60Hz~100Hz. The power drive has 7 outputs, which is adjustable to fit the external environment. Among them, sub-field scanning working mode is adopted in the design, gray scale is selectable.

Research on Structure of Communication Network in Smart Grid

As the next generation of power systems, smart grid is a high degree integration of electric power, communication and automatic control. A safe, effective and intelligence communication platform is the precondition of building Smart grid. This paper introduces an overview of power communication in nowadays, analyzes and probes the structure and safe strategy of communication network in Smart grid. Several key technologies of communication safety in Smart grid are summarized and generalized. This paper has important reference value to the research area of the communication safety of Smart grid.

Intelligent Remote Wireless Streetlight Monitoring System Based on GPRS

According to the development trend of the streetlight, this paper presents a remote and wireless streetlight monitoring system based on GPRS. GPRS stands for General Packet Radio Service, it has a lot of advantages like widely used, high transmission speed, low power consumption and so on. The system uses microcontroller chip MSP430F2274 and wireless transceiver chip CC2500. The whole network consists of the control center and up to 100 groups of control network, and each control network has up to 100 terminal nodes, terminal nodes measure humidity, current, voltage and other information, and send these information to the transition points by the RF wireless transceiver module, then the transition points transmit information to the control center through GSM/GPRS networks. The control center will deal with the data so that it can know the situation of each streetlight. According to the result the control center gives orders to each streetlight to control the switch state and illumination of them.