Search

This paper mainly studies the tracking algorithm on intelligent vehicle based on image recognition. The intelligent vehicle system is designed with two parts: the hardware part, which is composed of drive motor, steering gear, rechargeable battery, image sensor and control circuit; the software part, which is designed in development environment and run in ARM processor. A fully working prototype system has been developed, test results show that the fastest time which refers to the intelligent vehicle running on the special track is 19 seconds, and it can automatically finish navigating with the camera, the result is good. The intelligent vehicle system was used to participate in the National mobile robot competition and won the National Award.

Refueling robot to complete the refueling can effectively improve gas station efficiency, reduce pollution and save labor costs. In this paper, a brief description of the main functional sequence in fueling robots is given. The design adopts the machine vision technology to locate the fuel tank port intelligently. Based on binocular vision-based positioning and recognition strategy, two cameras are used to capture the target image, the feature points are matched by SIFT algorithm, and the image coordinates of the same three-dimensional point on different images are obtained by wavelet mutual information registration. Finally, the corresponding position and posture parameters are identified and the fuel cap is positioned, which provides the parameter basis for the driving and controlling of the refueling robot.

The image super-resolution reconstruction can be widely used in military, medicine, transportation, etc. Due to the unfavorable factors, such as noises and obstacles, of the low resolution images, the influence of noises or the obstacles can be amplified by the super-resolution reconstruction. For this reason, the wavelet analysis were combined with the particle swarm optimization algorithm in this paper. Through the wavelet and particle swarm de-noising and registration on the original low resolution images, the super resolution reconstruction image can be obtained after the image fusion, and the results are better than traditional methods.

The super-resolution reconstruction is an important method for improving the image quality, and the objective world can be better understood from 3D super-resolution model than from 2D images. In this paper, the image super-resolution reconstruction algorithm based on wavelet and the 3D geometric modeling technology were combined to study the 3D super-resolution reconstruction of images. Firstly, the different super-resolution images of different angles were obtained by using the wavelet super-resolution reconstruction. Secondly, the 3D information was extracted by adopting the three-dimensional voxel model. Finally, the super-resolution image data about the real objects and the super-resolution 3D geometric model were synthetically acquired, which can lay the theoretical basis for further image processing and model development.

The super-resolution reconstruction is an important method for improving the image quality, which possesses widely application in the fields of video monitoring, remote sensing imaging, medical image processing, etc. In this paper, firstly, the salient edges and the boundaries of salient objects were obtained by using the salient characteristics of visual attention based on computer visual model and wavelet analysis; secondly, the super-resolution image was synthetically acquired by adopting the bilinear interpolation in the contour's internal region and the adaptive interpolation in the boundaries' external region respectively. The simulation results showed that the visual effect of the reconstructed model was more according with human visual characteristics. In addition, the subjective judgment and objective evaluation is better than traditional reconstruction algorithm. The algorithm in this paper achieves good effects and reaches good feasibility and validity.

Support vector machine is a machine learning method. It takes on the good generalization ability and prediction accuracy. But the parameters of SVM model seriously affect the generalization ability and prediction accuracy of SVM model on the great extent. So an improved particle swarm optimization (PSO) algorithm based on chaotic search is introduced into the SVM model in propose a novel data classification (AMPSVM) method for processing the complex data. The first, the ergodicity, stochastic property, and regularity of chaos is used to chaotically search the current best individual, which randomly replaces the selected individual in the population in order to speed up evolution, improve the searching ability, convergence speed and accuracy. Then the improved PSO algorithm is used to select and optimize the parameters of the SVM (AMPSVM) model in order to improve the learning performance and generalization ability of the SVM model. In order to verify the effectiveness of the AMPSVM method, UCI data is selected in here. The experiment results show that the proposed AMPSVM method takes on the strong generalization ability, best sensitivity and higher classification accuracy.

Image super-resolution reconstruction was an important task in the field of image enhancement and image restoration. In this paper, a new super-resolution reconstruction method which based on wavelet analysis and sub-pixel interpolation was proposed by using wavelet edge detection and polynomial subdivision location. The simulation results showed that the boundary of reconstructed high-resolution image is clear and natural, and the subjective judgment and objective evaluation is better than traditional reconstruction algorithm. The algorithm in this paper achieves good effects and reaches good feasibility and validity.

Video face recognition has become one of the most active research direction in the field of face recognition. In this paper, the influence of the illumination and the posture was analyzed and the corresponding solution was designed. Through the combined face recognition algorithm of global features and local features based on the wavelet analysis, as well as the Hidden Markov model using as the post-processing, the conception of optimizing recognition was set up for construction. The feasibility of the model was proved by the partial experiments, which provide the beneficial reference for intelligent monitoring technology.

There are some problems in the video identity recognition, such as low efficiency of data processing and instability of the identification performance. This paper designs specific links and the corresponding processes for the interested character recognition and scene tracking, which are including the construction of adaptive segmentation, video face flexible recognition, feature library model, optimizing search strategies based on genetic algorithm, scene path adaptive generation, all these are in order to provide reference for automatic interested character recognition and scene tracking.