Xiong Youlun

High Probability Impulse Noise-Removing Algorithm Based on Mathematical Morphology

A novel filter based on mathematical morphology for high probability impulse noise removal is presented. First, an impulse noise detector using mathematical residues is proposed to identify pixels that are contaminated by the salt or pepper noise. Then the image is restored using specialized open-close sequence algorithms that apply only to the noisy pixels. Finally, black and white blocks that degrade the quality of the image will be recovered by a block smart erase method. Experimental results demonstrate that the proposed filter outperforms a number of existing algorithms and is particularly effective for the very highly corrupted images

General Criterion And Control Strategy Of Collision-free Movement For Manipulators

In this paper, we present a general procedure for finding collision-free paths for a polyhedral object moving through space that is cluttered with obstacle polyhedra. The unified criterion of collision- free motion is represented as Conv {H/sub i/)Conv {O/sub j/)=/spl Oslash/ which is suitable for all cases regardless of the space dimensionality. Based on the criterion, we are led to construct the describing function J which stands for minimal distance between the moving object and obstacles. Tha algorithm derived from describing function J can be used to find any path of interest for both the 2-D problem and the 3-D problem. It is shown that the solution method to the find path problem may be reduced to a search technique. Since the describing function J is related to only some extreme points, the strategy for automatically determining safe paths between configurations in presence of obstacles may be effectively sets implemented by a simple programming.

QTS ALOHA: A Hybrid Collision Resolution Protocol for Dense RFID Networks

The performances of the frame slotted ALOHA based and tree based protocols are unsatisfactory in resolving the collision occurring in dense RFID networks made up of hundreds or even thousands RFID tags. QTS ALOHA, a hybrid collision resolution protocol which combines both the query tree protocol and the dynamic frame slotted ALOHA protocols, is proposed to achieve the rapid identification of a large amount memory-less RFID tags. For the identification of a group of tags, based on the data already collected, the population of tags in the group is estimated, and according to the identification accuracy required by the application system, the group of tags may be split into multiple subgroups through some binary suffix strings, dynamic slotted ALOHA protocol is performed to identify tags and collect data in the group or subgroups. This collision resolution process is performed repeatedly until the required identification accuracy is achieved. Numeric simulation verifies that this hybrid protocol outperforms othr protocols in frames and slots consumed to identify different number of tags, keeps a high throughput in each identification frame, and is especially suitable for the collision resolution in dense memory-less passive RFID networks.

Virtual manufacturing systems and environment

Virtual manufacturing (VM) is a simulation method whereby it is possible to estimate manufacturing processes without using real facilities. This paper gives an overview of the modelling and simulation environment of virtual manufacturing systems, and introduces the technologies of modelling and simulation such as object-oriented technology. We also discussed the virtual manufacturing systems framework, virtual manufacturing environment, and the automated manufacturability analysis. The main research issues of virtual manufacturing are discussed.

Edge Detection of the Low Contrast Welded Joint Image Corrupted by Noise

To extract the edges from the low-contrast welded joint image corrupted by noise, the morphological filter is firstly proposed to remove the noise from the image. Then the improved fuzzy edge detection (FED) algorithm is adopted to realize the multi-level fuzzy enhancement of the filtered image and extract the edges from the enhanced image using multi-directional edge detection operator. Experimental results demonstrate that the improved FED algorithm can extract the continuous and thin edges and remove the false edges from the image, which leads to its better performance than the Sobel operator, Canny operator, traditional FED algorithm and fast FED algorithm.

A comparative study on the splitting tree based protocols for RFID tag identification

Collision caused by that multiple RFID tags try to transmit their identifiers through the shared wireless communication channel simultaneously is one of the key issues that affect the ubiquitous adoption of RFID system. In this paper, based on the stochastic distribution model, a simple but efficient tag population estimation method is suggested to work in the splitting tree collision resolution protocol, and the estimation accuracy is also discussed. Based on tag population estimation, the group of RFID tags to be identified is divided into multiple subgroups to be resolved in following slots. This group splitting process is performed recursively until that in each subgroup, there is zero or one tag, so all subgroups can be resolved and all tags can be identified. The performances of different splitting tree protocols are evaluated and some useful conclusions are achieved.

A SAT Method for Improving Test Pattern Generation

This paper presents a solution for test pattern generation (TPG) based on Boolean satisfiability (SAT). The key to a SAT-solver can be scalable is that it is able to take into account the information about high-level structure of formulas. The paper augments a circuit structure layer to the SAT-solver to maintain circuit-related information and value justification relations. It dovetails binary decision graphs (BDD) and SAT techniques to improve the efficiency of test pattern generation. More specifically, it first exploits inexpensive reconvergent fanout analysis of circuit to gather information on the local signal correlation by using BDD learning. It then uses the above learned information to restrict and focus the overall search space of SAT-based TPG. Its learning technique is effective and lightweight. The experimental results demonstrate the effectiveness of the approach.

An overview on distance and pseudo distance functions and their applications

The fast and accurate collision detection and distance calculation between geometric models is fundamental in robotics, manufacturing and computer-simulated environments. This paper surveys the state of the art in the theory and algorithms of distance and pseudo distance functions between convex sets. An overview on the existing results, including fast algorithms for distance calculation, definitions and properties of various pseudo distance functions (the J-function, the growth distance, and the pseudo minimum translational distance), is presented. The distance and pseudo distance functions are versatile in a wide range of areas. Their applications in robot motion planning, grasp analysis/synthesis, form error evaluation, and engineering optimization is demonstrated.

Stochastical Model and Performance Analysis of Frequency Radio Identification

RFID (Radio Frequency Identification) can assign a unique digital identifier to each physical item, and provide an efficient, cheap and contactless method for gathering the information of the physical items to enable their automatic tracking and tracing(Finkenzeller 2003). RFID technology serves as the back stone of the “Internet of Things”(Engels 2001), and is reviewed as a main enabler of the upcoming “Pervasive Computing”(Stanford 2003). RFID systems have been widely adopted in quite a lot applications, ranged from “smart box” to world-wide logistics management systems. A typical RFID system is consisted of some RFID tags, one or more RFID readers and the backend information system. Each RFID tag holds a unique identifier and is attached to a physical item. RFID reader is used to collect the identifiers stored in the RFID tags located in its vicinity and is often connected with the backend information system. During identification, The RFID reader asks the RFID tags to modulate their binary identifiers into signals and transmit these signals back to the reader through the air interface, which is a wireless communication channel for the RFID tag and reader to exchange information. Afterwards, The RFID reader sends the data gathered to the backend information system for further processing and dispatching to various applications. One of the main issues that affect the universal deployment and application of the RFID system is the collisions occurred during RFID tag identification(Wu 2006). The simultaneous modulated signals broadcasted by the RFID reader and transmitted from the RFID tags will interfere in the air interface, in which case, what the receivers can get is only a collision signal but no useful information. Collisions occurred in the RFID system can be categorized as the reader-reader, reader-tag and tag-tag collisions(Shih 2006). When two or more RFID readers try to broadcast messages through the air interface simultaneously, reader-reader collision occurs. Due to that they are often connected with the computer system and can be equipped with enough resource to monitor the air interface, RFID readers can detect the collision and coordinate with each other in advance. Reader-reader collision can be avoided and resolved completely with some deliberate designed protocols, such as the ColorWav(Waldrop 2003) and others(Leong 2006).

A Hybrid Collision Resolution Protocol for Passive RFID Tag Identification

Wireless communication channel collision of RFID system is a key issue that affects the pervasive application of the system, especially for passive RFID system due to the constrained communication ability and energy supply of tags. The collision resolution protocols suggested can be categorized into deterministic tree based protocols and probabilistic frame slot ALOHA based protocols. A hybrid protocol, which combines the query tree protocol and dynamic framed slotted ALOHA protocol, is presented here. In this protocol, at the beginning of each identification frame, the reader broadcasts a query command message which contains a suffix string and integer to indicate a frame size used in this frame to all tags in its vicinity, and every tag whose data match the specified suffix string randomly selects a slot in the frame to transmit its data to the reader. After each round, based on the different occupying situation of slots, the reader re-estimates the number of tags in its range, and according to the identification accuracy required, decides whether to split the group of tags to be identified into sub-groups or not, and adjust the suffix string and select a suitable frame size for the command messages of the following frames. Through this repeated identification and division process, the RFID tags can be identified within a number of frames. Some key factors that influence the efficiency of the identification are also discussed based on binomial distribution model. Three methods for the tag population estimation are examined and their accuracies are compared, and the optimum frame size for the identification of different number of tags is given in consideration of the throughput and efficiency. To verify the performance of the hybrid protocol, a numeric simulation is performed, in which 100 groups of SGTIN-96 encoded tags are generated randomly. Simulation shows that this hybrid protocol outperforms other frame slot ALOHA based collision resolution protocols, especially when there are a large number of RFID tags to be identified. The overall throughput of the hybrid protocol is kept around 34.8% when there are more than 40 tags, near the theoretical maximum value 36.8%. A linear relationship between the number of tags and the total frames and slots required is also observed in the numeric simulation.