Yu Shuchun

Preprocessing for stereo vision based on LOG filter

In order to improve LOG filter performance for stereo vision, a new preprocessing method was proposed. This method was divided into 4 steps: firstly wipe out random noise with a mean filter, then reduce Gaussian noise with a Gaussian filter, and next balance brightness difference between stereo image pairs with the approach histogram equalization, and finally enhance image edges and details with a Laplacian operator. Experimental results demonstrate that this preprocessing method can eliminate image noise and reduce brightness difference effectively.

Image Retrieval Based on Multi-feature Fusion

In content-based image retrieval, and for this critical issue of image feature fusion, paper proposes a new method to determine the weights for multi-feature fusion. In this paper, color histogram, color correlogram, gray level co-occurrence matrix, Tamura and Hu moments, this five kinds of feature extraction method was adopted. Firstly, use these five features conducted single feature retrieval on the various types of images to determine the precision rate of each feature retrieval and compare their precision rate. Through precision rate to determine the dynamic weight of various features when conducting the feature fusion retrieval in different categories images. The experimental results showed that: according the precision rate of each feature to dynamically regulate the weights, when carrying multi-feature fusion retrieval for different types of image, compared to multi-feature retrieval with fixed weights, precision rate of retrieval has improved significantly.

Video inpainting model for camera motion based on improved background subtraction method

Motion object detection has been widely used in traffic monitoring and target tracking fields. In order to solve the difficulty of building a model of the background and improving the accuracy of the update rate in the background subtraction, a new detection moving object method combining the surf algorithm and the background subtraction is proposed in this paper. Firstly, according to the continuity of motion, motion regions are labeled and filled in; Secondly, the image mosaic by surf algorithm is conducted to obtain the image with the whole background, then the whole background image mosaic is used to obtain the frame image with the whole background; Finally, frame difference method is used to obtain the foreground object, and then a morphology is processed. The experiment has not only achieved higher accuracy which is improved by about 8%~10%, but also got lower error which is reduced by 3.5%~4.0%. In a word, the algorithm has better robustness.

Research and achievement of QR code encryption based on cellular automata

To improve the security performance of the information stored in QR (Quick Response) Code, a QR code encryption and decryption method based on elementary cellular automata state rings was proposed in this paper. The cellular automata can simulate complex phenomenon just using simple dynamical system. Based on this feature, the method uses the cellular automata to encrypt and decrypt QR code binary image with such parameters: length is 8, the cyclic boundary conditions and the state space of {0, 1}. The experimental results show that the method proposed in this paper has some advantages, such as high speed, good effect and high security.

A novel motion object detection method based on improved frame difference and improved Gaussian mixture model

The existing motion detection methods include background subtraction and frame difference. But it is prone to exist some holes with frame difference method and it is difficult to build background model using background subtraction method. So the test results did not achieve the ideal state. Aim at these problem, this paper combines frame difference method improved by motion history image with background subtraction method based on improved Gaussian mixture model to detect the motion object. The experimental results show the method has achieved a satisfactory effect.