Minhua Li

Shandong University of Science and Technology, Tsingtao, Shandong Sheng, China

Are you Minhua Li?

Claim your profile

Publications (9)0 Total impact

  • Minhua Li, Meng Bai
    [Show abstract] [Hide abstract]
    ABSTRACT: Text contained in images and video frames provide an important clue for content based images and video indexing. However, it is difficult to segment text from images with complex background. This paper proposes a new proposal for text segmentation from images with complex background based on Markov random field. Experimental results demonstrate the performance of the proposed method.
    01/2012;
  • Minhua Li, Meng Bai
    [Show abstract] [Hide abstract]
    ABSTRACT: A novel text detection method based on multi-structure elements of morphology is presented to extract text lines in this paper. Firstly, a multi-structure elements of morphology-based scheme is executed to extract contrast edges from color images. Then connected component analysis and a cascade classification scheme are adopted to obtain candidate text lines. Finally, a projection based method is applied to extract character characteristics to filter out impossible text lines. Experimental results indicate that the proposed method can effectively detect text lines in different font-size, language and background complexity.
    Industrial Control and Electronics Engineering (ICICEE), 2012 International Conference on; 01/2012
  • Minhua Li, Meng Bai
    [Show abstract] [Hide abstract]
    ABSTRACT: To detect text from an image with a different background, an adaptive text detection method based on image complexity analysis is proposed. Before text detection, this approach adopts an image complexity analysis step to classify image complexity into three categories: low complexity, middle complexity and high complexity. Then images with different complexity adopt different methods to extract image edges. The proposed text detection method takes a coarse to fine detection strategy which combines the edge-based method, connected component based method and the texture based method into a framework. Experimental results demonstrate the performance of the proposed method.
    Intelligent Control and Automation (WCICA), 2012 10th World Congress on; 01/2012
  • Minhua Li, Meng Bai
    [Show abstract] [Hide abstract]
    ABSTRACT: Yaw dynamic model of a miniature unmanned helicopter is needed to develop for heading control. A yaw dynamic model is deduced based on miniature unmanned helicopter characteristics in hover. Different from a large helicopter, the yaw damping system of a miniature helicopter is realized through the negative feedback of helicopter heading rate, which is provided by an angular rate gyro. Akaike Information Criterion is used to solve the problem of determining model order. Based on flight experimental data, a least square method is adopted to estimate the unknown parameters in the yaw dynamic model. And the identified model is verified by comparing the model output data with the collected flight experiment data.
    Intelligent Control and Automation (WCICA), 2012 10th World Congress on; 01/2012
  • Meng Bai, Minhua Li
    [Show abstract] [Hide abstract]
    ABSTRACT: For SINS/GPS integrated navigation system with unknown measurement noise covariance matrix, adopting the conventional Kalman filtering approach to estimate the navigation system errors will lead to a large state estimation error or even make the filter diverge. To solve this problem, an adaptive sequential Kalman filter is presented, in which the measurement noise covariance matrix is estimated on-line by an innovation-based adaptive estimation (IAE) method. Properly designed discontinuous feedback control law and serial measurement processing make the adaptive filter more suitable for real time implementation. Simulation results reveal that without an exact measurement noise covariance matrix, the adaptive sequential Kalman filtering approach can still estimate the errors of SINS/GPS integrated navigation system effectively.
    Intelligent Control and Automation (WCICA), 2012 10th World Congress on; 01/2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: Text contained in images and video frames provide important clues for information indexing and retrieval. But it is difficult to segment text from images, especially those images with complex background. This paper presents a new conditional random field approach, in which contextual features are introduced into text segmentation. Local visual information and contextual label information are integrated into a conditional random field by several components. Some components focus on visual image information to predict the category within the image sites, while others focus on contextual label information to determine the patterns within the label field. Integrating contextual label information in conditional random field can effectively resolve local ambiguities and improve text segmentation performance in complex background. The comparing results demonstrate that the proposed method outperforms other methods for text segmentation from complex background.
    Pattern Recognition Letters. 01/2010;
  • Minhua Li, Chunheng Wang, Ruwei Dai
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we propose a new approach for unconstrained handwritten character recognition based on wavelet energy density feature (WEDF) and multilayer neural network. Unlike other method taking the wavelet coefficients directly as features, our method using the wavelet energy density features instead. The proposed approach consists of a feature extraction stage for extracting wavelet energy density features with wavelets transform, and a classification stage for classifying handwritten characters with a simple neural network. In order to verify the performance of the proposed method, experiments are carried out on handwritten numerals recognition. Experimental results indicate that the WEDF is stable and reliable in handwritten character recognition and performs better than wavelet coefficient feature, it provides high recognition rate on both training samples and testing samples.
    Intelligent Control and Automation, 2008. WCICA 2008. 7th World Congress on; 07/2008
  • Minhua Li, Chunheng Wang
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, an adaptive edge-based text detection approach in images and video frames is proposed. The proposed approach can adopt different edge detection methods according to the image background complexity. It mainly consists of four stages: Firstly, images are classified into different background complexities. Secondly, different edge detectors are applied on the images according to their background complexities. Thirdly, connected component analysis is adopted on the edge image to obtain text candidates. Finally, the text candidates undergo the refinement algorithm to find the exact position. Experimental results demonstrate that the proposed approach is robust to text size and could effectively detect text lines in images and video frames in both simple background and complex background.
    Neural Networks, 2008. IJCNN 2008. (IEEE World Congress on Computational Intelligence). IEEE International Joint Conference on; 07/2008
  • Minhua Li, Ruwei Dai, Yaodong Li
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we put forward a novel human computer cooperation system the scheme producing and evaluation system for the hall for workshop of metasynthetic engineering (HWME).The proposed system can produce schemes automatically and evaluate schemes effectively whose functions are implemented by the cooperation of human and computers. The system mainly consists of three parts: the scheme framework model, the scheme forming model and the scheme evaluation model. Firstly, the scheme framework model is adopted for group experts to produce a scheme framework to write scheme contents; secondly, the scheme forming model is applied for group experts to write scheme contents and then the computer combine the similar contents to form optional schemes; finally, the scheme evaluation model is introduced to evaluate the optional schemes to obtain the optimal scheme. Experimental results demonstrate that the proposed system is feasible and effective for experts in HWME to solve complex problem which provides a good tool for experts to make decisions and schemes.
    Proceedings of the 32nd Annual IEEE International Computer Software and Applications Conference, COMPSAC 2008, 28 July - 1 August 2008, Turku, Finland; 01/2008

Publication Stats

10 Citations

Top Journals

Institutions

  • 2010–2012
    • Shandong University of Science and Technology
      Tsingtao, Shandong Sheng, China
  • 2008–2010
    • Northeast Institute of Geography and Agroecology
      • Laboratory of Complex Systems and Intelligence Sciences
      Beijing, Beijing Shi, China