Lingqiu Zeng

Automata-based L-Grammar extraction from multiple images for virtual plants

L-system (Lindenmayer system) and its application have been one of the most famous and powerful tools for virtual plant modelling. But it is really hard to develop L-grammar manually for a given plant depending only on imagination or experience. For bridging this gap, a novel automatic L-grammar extraction approach is presented in this work. Initially, image processing as well as pattern recognition methods are employed to recover morphological and geometrical information for growth units and metamers. And then, these data are further analyzed using Markovian methods and acted as parameters for bidimensional hierarchical automata (BHA) to describe plant branching structure. Finally, the L-grammar has been extracted by means of the transformation from BHA to L-system. Experimental results show that our approach can extract L grammar for unfoliaged tree effectively.

A Road Hotspots Identification Method Based on Natural Nearest Neighbor Clustering

During the last decade, the concept of cluster, has become a popular practice in the field of road safety, mainly for the identification of worst performing areas or time slots also known as hotspots. However, current clustering methods used to identify road accident hotspots suffer from various deficiencies at both theoretical and operational level, these include parameter sensitivity, identify difficultly on arbitrary shape, and cluster numbers rationality. The objective of this study is to contribute to the ongoing research effort on hotspots identification. Employing the concept of natural neighbor, a new algorithm, named distance threshold based on natural nearest neighbor (DTH3N), is proposed in this paper, striving to minimize the aforementioned deficiencies of the current approaches. Experiment results show that, comparing with existing methods, proposed algorithm presents a better performance on cluster division. Furthermore, this new method can be viewed as an intelligent decision support basis for road safety performance evaluation, in order to prioritize interventions for road safety improvement.

Abnormal hotspots detection method based on region real-time congestion factor

Road hotspots detection method is a key issue in the field of intelligent transportation research. Compared with normal hotspots caused by high traffic flow, abnormal hotspots, which are results of road accidents, perform an occurrence time random behavior and difficult to predict. In this paper, a region real-time congestion factor is constructed to realize road abnormal hotspots discovery. Based on taxis GPS data of Hangzhou City, China, we analyze the relationship between proposed congestion factor and the real-time traffic data. Two accidental scenarios are built to verify the validity of the proposed method. The experiment results show that the proposed method performs well in real-time abnormal hotspot detection and analysis output could be useful in path planning and traffic management.

A large vehicle first clustering method based road section risk level estimation

On-road large vehicles are always considered as potential danger, which could generate serious damage and strongly threaten road safety. Hence, large vehicle related accident forecasting becomes an important research topic in intelligence transportation area. In this paper, a large vehicle first clustering method (LVFC) is proposed to estimate the risk level of vehicle group. Composite performance index (CSPI), which is used to illustrate the risk level of hotspots area, has become a popular practice in the field of road safety. CSPI value is employed to illustrate environment condition of object road section. Combining CSPI with vehicle groups risk level, a comprehensive risk parameter is presented to represent real-time transportation condition, which illustrates the risk level of a road section with large vehicle group passing by. A vehicle following model is designed to simulate the generation of large vehicle first clusters and its changing procedure, which provides justification for the proposed algorithm.

Leaf mutiscale variation algorithm under feedback control system

A new algorithm using Functional structural plant models (FSPM) employing feedback control system(FCS)to realize the multiscale change of the physiological parameter for leaf growth is proposed in this paper. Firstly, the plant branching structures are generated by Bidimensional Hierarchical Automata (BHA) while the disturbance function is used to realize the interaction between physiological parameter and botanic growth stimulant. Secondly, the growth stimulant changes the organic details and influences the environment parameters. On the other hand, the change of environment parameters modifies the plant branching structure in return. Thirdly, the varying vein texture is synthesized by reaction-diffusion principle based on the canalization hypothesis. Finally, the three-dimension deformation of a leaf is proposed by the controllability grid bending. Simulation results show that proposed algorithm can effectively simulate the varying process of leaf texture and form by changing physiological parameter. It can well meet the requirement of dynamic displaying plant organ in virtual agricultural laboratory.