Search

Vision-based detection of vehicle line-pressing violation has been widely used in intelligent transportation system. However, the accuracy of most current methods is low due to weather, illumination, shooting angle, etc., which increases the workload of manual verification. To cope with these problems, an efficient automatic verification method of vehicle line-pressing violation is proposed based on convolutional neural network (CNN) and geometric projection. First, a CNN model is used to detect vehicles and features are extracted to match the vehicles with line-pressing violation. Second, based on driving direction and vehicle 2D bounding box, pose relationship of the vehicles to the ground is fitted. Finally, according to the relationship between lane line and vehicle chassis range, automatic verification of vehicle line-pressing violation is realized. In the experimental results, the recall rate and accuracy rate reached 90.3 and 99.7% respectively in 200 test samples, which is better than those of current methods. The experimental results show that the proposed method has high accuracy, recall rate and good performance, and can be used for the automatic verification of vehicle line-pressing violation.

Tire burst of vehicles is one of the main causes of highway accidents. Furthermore, tire aging is the most important cause of tire burst. It will be of great significance to traffic safety, through identifying the tire text code (TTC) to determine whether or not a tire is over the service life. In this article, a coarse-to-fine method of detection and rectification for TTC is proposed. First, text regions are located using the deep convolutional neural network (DCNN) model. Second, through fitting the tire circle curve, the close-up image of TTC is rotated to a positive direction. Third, a text boundary points’ generation strategy is proposed, which is based on the rotating bounding box of each character. Fourth, the curved text is rectified and input to the DCNN recognition model. On the tire image dataset, the detection accuracy of TTC and characters is 94% and 96.5% respectively, and the recognition accuracy is 94.5%. Through additional experimental results, it is proven that the proposed method has strong robustness to illumination and camera shooting angle, which can meet the real-time detection and recognition requirements as well.

114 p. Graph-based semi-supervised learning have attracted large numbers of researchers and it is an important part of semi-supervised learning. Graph construction and semi-supervised embedding are two main steps in graph-based semi-supervised learning algorithms. In this thesis, we proposed two graph construction algorithms and two semi-supervised embedding algorithms. The main work of this thesis is summarized as follows:1. A new graph construction algorithm named Graph construction based on self-representativeness and Laplacian smoothness (SRLS) and several variants are proposed. Researches show that the coefficients obtained by data representation algorithms reflect the similarity between data samples and can be considered as a measurement of the similarity. This kind of measurement can be used for the weights of the edges between data samples in graph construction. Each column of the coefficient matrix obtained by data self-representation algorithms can be regarded as a new representation of original data. The new representations should have common features as the original data samples. Thus, if two data samples are close to each other in the original space, the corresponding representations should be highly similar. This constraint is called Laplacian smoothness.SRLS graph is based on l2-norm minimized data self-representation and Laplacian smoothness. Since the representation matrix obtained by l2 minimization is dense, a two phrase SRLS method (TPSRLS) is proposed to increase the sparsity of graph matrix. By extending the linear space to Hilbert space, two kernelized versions of SRLS are proposed. Besides, a direct solution to kernelized SRLS algorithm is also introduced.2. A new sparse graph construction algorithm named Sparse graph with Laplacian smoothness (SGLS) and several variants are proposed. SGLS graph algorithm is based on sparse representation and use Laplacian smoothness as a constraint (SGLS). A kernelized version of the SGLS algorithm and a direct solution to kernelized SGLS algorithm are also proposed. 3. SPP is a successful unsupervised learning method. To extend SPP to a semi-supervised embedding method, we introduce the idea of in-class constraints in CGE into SPP and propose a new semi-supervised method for data embedding named Constrained Sparsity Preserving Embedding (CSPE).4. The weakness of CSPE is that it cannot handle the new coming samples which means a cascade regression should be performed after the non-linear mapping is obtained by CSPE over the whole training samples. Inspired by FME, we add a regression term in the objective function to obtain an approximate linear projection simultaneously when non-linear embedding is estimated and proposed Flexible Constrained Sparsity Preserving Embedding (FCSPE).Extensive experiments on several datasets (including facial images, handwriting digits images and objects images) prove that the proposed algorithms can improve the state-of-the-art results.