Aiming at the problem that the characteristic and evaluation indexes for reasearch of traffic behavior spectrum are incompleted both at home and abroad and quantitative analysis cannot be performed in the research, the corresponding characteristic and evaluation indexes were defined to establish a complete traffic behavior spectrum system with quantitative analysis of regional traffic behavior data. Firstly, based on the characteristics of traffic behavior, an improved Analytic Hierarchy Process (AHP) was used to classify the traffic order types. Secondly, Real-Time System Integration (RTSI) algorithm with multi-data fusion was used to comprehensively evaluate the traffic safety of a certain road. Finally, a traffic behavior spectrum analysis tool was developed, calculating traffic safety index of a road section according to the traffic live data, and analyzing traffic behavior in the section more completely.

Aiming at the main problem of the Sparse Representation (SR) coefficients of the image blocks in image super-resolution method, an enhanced self-learning super-resolution approach for single image was proposed by using the weighting idea. Firstly, the pyramid of high and low resolution images was established by self-learning. Then, the image block feature of low-resolution images and the central pixels of the corresponding high-resolution image blocks were extracted respectively. The role of the center pixel in constructing the image block sparse coefficient was emphasized by giving different weights of different pixels in the image blocks. Finally, the combination of SR theory and Support Vector Regression (SVR) technique was used to build the super-resolution image reconstruction model. The experimental results show that compared with the Self-Learning Super-Resolution method for single image (SLSR), the Peak Signal-to-Noise Ratio (PSNR) of the proposed method is increased by an average of 0.39 dB, the BRISQUE (Blind/Reference-less Image Spatial Quality Evaluator) score of no-reference image quality evaluation criteria is reduced by an average of 9.7. From the subjective perspective and objective values, it is proved that the proposed super resolution method is more effective.

The acoustic signal waveform of vehicle through the deceleration zone is different from that of the normal running on the road, the extraction of its feature parameters is crucial to the automatic judgment of number, speed and type of vehicles, and the acoustic signal envelope curve has many advantages in extracting its feature parameters compared with the original signal. However, the traditional envelope extraction algorithm has the problems that there are many burrs and it is difficult for the feature parameters to truly reflect the signal properties and features in the envelope extraction of acoustic signals in such traffic domain. In order to solve the problems, combined with the characteristics of acoustic signals of vehicles through the deceleration zone, a new envelope extraction algorithm for acoustic signal of vehicle pressing line based on variable step size was proposed. Different step sizes were set to traverse the signal. The curve was plotted by using the maximum point in each step and compared with the original signal waveform. The sharp definition and error of feature point extraction were taken as the judgment basis to realize the effective extraction of the acoustic signal envelope. The experimental results show that, under the same number of sampling points, the extracted envelope curve by the proposed algorithm is more clear and has less burrs than that by the traditional envelope extraction algorithm, and the extraction error of feature parameters is smaller.

Some problems exist in the traditional typhoon disasters prediction, such as depending mainly on statistical methods, lack of semantic-driven processes and intelligent reasoning. In order to solve these problems, a reasoning mechanism of the ontology model of typhoon disasters domain based on Jena was proposed. Firstly, the ontology model of typhoon disasters domain expressed by Web Ontology Language (OWL) was built based on analyzing the impact factors and evolution history of the typhoon disasters fully. Secondly, the Jena reasoning engine and custom rules were used for the reasoning of typhoon disasters ontology model and mining the hidden impact factors of typhoon disasters or the information of disaster chains. Finally, an Ontology-driven Typhoon Disasters Expert System (Onto-TDES) was built. The experimental results show that the mechanism can solve the lack of semantic-driven processes and intelligent reasoning preliminarily, and improve the intelligent level of the management and prediction of the typhoon disasters.

Traditional wavelet-based algorithm has a common effect on the images of light nonuniformity and scarcity. Aiming at the shortcoming, a new wavelet-based algorithm for image enhancement was proposed. The image was first decomposed into multi-level wavelet to obtain the scaling coefficients and the multi-level wavelet coefficients. Then, every level of wavelet coefficients was enhanced by different algorithms, and the scaling coefficients were processed by MSR(Multiscale Retinex). Finally, the image of enhancement was obtained via the inverse wavelet transform. Experiments show that the algorithm excels conventional algorithms in the effect of enhancement and the abatement of noise, at the same time, it has an excellent effect on the images of light nonuniformity and scarcity.