Search Results

You are looking at 1 - 4 of 4 items for :

  • Author: Jian Su x
  • Control Engineering, Metrology and Testing x
Clear All Modify Search
Open access

Guan Xu, Xiaotao Li, Jian Su, Rong Chen and Jianfang Liu

A Method for the Estimation of the Square Size in the Chessboard Image using Gray-level Co-occurrence Matrix

The paper proposes a new simple procedure for measuring the square size employing the gray-level co-occurrence matrix of a chessboard image. As the size of the square structure in a chessboard image provides the geometric constraint information among the corners, it is available to improve the precision of extracting corners and serve the camera calibration. The co-occurrence matrix of a chessboard image is constructed to obtain the statistic information of the grayscale distribution. The 2D offset of the matrix is parameterized to calculate the correlation which is regarded as the implication of the repetition probability of the similar textures. A descending tendency is observed in the experiments because the similarity decreases with the greater offset. However, minimum and maximum are captured in the correlation curve, which represents that the square texture reappears with the periods of one and two square size, separately. The size of the square is tested by applying the first minimum of the correlation. The experiments are performed on the horizontal and vertical directions which are corresponding to the length and the width of the square, respectively. The experiments prove that the described method has the potential to measure square size of the chessboard.

Open access

Dongliang Su, Jian Wu, Zhiming Cui, Victor S. Sheng and Shengrong Gong

This paper proposes a novel invariant local descriptor, a combination of gradient histograms with contrast intensity (CGCI), for image matching and object recognition. Considering the different contributions of sub-regions inside a local interest region to an interest point, we divide the local interest region around the interest point into two main sub-regions: an inner region and a peripheral region. Then we describe the divided regions with gradient histogram information for the inner region and contrast intensity information for the peripheral region respectively. The contrast intensity information is defined as intensity difference between an interest point and other pixels in the local region. Our experimental results demonstrate that the proposed descriptor performs better than SIFT and its variants PCA-SIFT and SURF with various optical and geometric transformations. It also has better matching efficiency than SIFT and its variants PCA-SIFT and SURF, and has the potential to be used in a variety of realtime applications.

Open access

Jian Wu, Zhiming Cui, Victor S. Sheng, Pengpeng Zhao, Dongliang Su and Shengrong Gong

SIFT is an image local feature description algorithm based on scale-space. Due to its strong matching ability, SIFT has many applications in different fields, such as image retrieval, image stitching, and machine vision. After SIFT was proposed, researchers have never stopped tuning it. The improved algorithms that have drawn a lot of attention are PCA-SIFT, GSIFT, CSIFT, SURF and ASIFT. In this paper, we first systematically analyze SIFT and its variants. Then, we evaluate their performance in different situations: scale change, rotation change, blur change, illumination change, and affine change. The experimental results show that each has its own advantages. SIFT and CSIFT perform the best under scale and rotation change. CSIFT improves SIFT under blur change and affine change, but not illumination change. GSIFT performs the best under blur change and illumination change. ASIFT performs the best under affine change. PCA-SIFT is always the second in different situations. SURF performs the worst in different situations, but runs the fastest.

Open access

Guan Xu, Xinyuan Zhang, Xiaotao Li, Jian Su and Zhaobing Hao


We present a reliable calibration method using the constraint of 2D projective lines and 3D world points to elaborate the accuracy of the camera calibration. Based on the relationship between the 3D points and the projective plane, the constraint equations of the transformation matrix are generated from the 3D points and 2D projective lines. The transformation matrix is solved by the singular value decomposition. The proposed method is compared with the point-based calibration to verify the measurement validity. The mean values of the root-mean-square errors using the proposed method are 7.69×10−4, 6.98×10−4, 2.29×10−4, and 1.09×10−3 while the ones of the original method are 8.10×10−4, 1.29×10−2, 2.58×10−2, and 8.12×10−3. Moreover, the average logarithmic errors of the calibration method are evaluated and compared with the former method in different Gaussian noises and projective lines. The variances of the average errors using the proposed method are 1.70×10−5, 1.39×10−4, 1.13×10−4, and 4.06×10−4, which indicates the stability and accuracy of the method.