Estimating fisheye camera parameters from homography

This paper presents a method to linearly estimate fisheye camera model parameters from the homography induced by the space plane between two fisheye images.The homography is firstly calculated by using four feature points,instead of three points on the same line,in the fisheye image.And the constraint on the model parameters of fisheye camera can be derived from the homography under the assumption that fisheye camera model is a polynomial model.Then the model parameters for different order polynomials are computed.The proposed technique requires only multiple fisheye images to include a planar scene and need not a priori knowledge of 3D coordinate of the planar scene.Experimental results with synthetic data and real fisheye images demonstrate the validity of our method.The method can also be extended to fisheye image of other planar scene excluding the planar calibration object.     

Fiducial Marker Based on Projective Invariant for Augmented Reality

Fiducial marker based Augmented Reality has many applications.So far the inner pattern of the fiducial marker is always used to encode the markers.Thus a large portion of the fiducial marker image is used for encoding instead of providing corresponding feature points for pose accuracy.This paper presents a novel method which utilizes directly the projective invariant contained in the positional relation of the corresponding feature points to encode the marker.The proposed method does not require the region of pattern image for encoding any more and can provide more corresponding feature points so that higher pose accuracy can be achieved easily.Many related approaches such as cmnulative distribution function,reprojection verification and robust process are proposed to overcome the problem of sensibility of the projective invariant.Experimental results show that the proposed fiducial marker system is reliable and robust,and can provide higher pose accuracy than that achieved by existing fiducial marker systems.     

Extracting 3D model feature lines based on conditional random fields

We propose a 3D model feature line extraction method using templates for guidance. The 3D model is first projected into a depth map, and a set of candidate feature points are extracted. Then, a conditional random fields (CRF) model is established to match the sketch points and the candidate feature points. Using sketch strokes, the candidate feature points can then be connected to obtain the feature lines, and using a CRF-matching model, the 2D image shape similarity features and 3D model geometric features can be effectively integrated. Finally, a relational metric based on shape and topological similarity is proposed to evaluate the matching results, and an iterative matching process is applied to obtain the globally optimized model feature lines. Experimental results showed that the proposed method can extract sound 3D model feature lines which correspond to the initial sketch template.     

Design of robotic visual servo control based on neural network and genetic algorithm

A new visual servo control scheme for a robotic manipulator is presented in this paper, where a back propagation (BP) neural network is used to make a direct transition from image feature to joint angles without requiring robot kinematics and camera calibration. To speed up the convergence and avoid local minimum of the neural network, this paper uses a genetic algorithm to find the optimal initial weights and thresholds and then uses the BP algorithm to train the neural network according to the data given. The proposed method can effectively combine the good global searching ability of genetic algorithms with the accurate local searching feature of BP neural network. The Simulink model for PUMA560 robot visual servo system based on the improved BP neural network is built with the Robotics Toolbox of Matlab. The simulation results indicate that the proposed method can accelerate convergence of the image errors and provide a simple and effective way of robot control.     

Self Localization Method Using Parallel Projection Model for Mobile Sensor in Navigation Applications

This paper presents a novel self localization method using parallel projection model for mobile sensor in navigation applications.The algorithm estimates the coordinate and the orientation of mobile sensor using projected references on visual image.The proposed method considers the lens non-linearity of the camera and compensates the distortion by using a calibration table.The method determines the coordinates and orientations with iterative process,which is very accurate with low computational demand.We...     

Tire Defect Detection Using Local and Global Features

In this paper, we present a tire defect detection algorithm based on sparse representation. The dictionary learned from reference images can efficiently represent the test image. As the representation coefficients of normal images have a specific distribution, the local feature can be estimate by comparing representation coefficient distribution. Meanwhile, a coding length is used to measure the global features of representation coefficients. The tire defect is located by both these local and global features. Experimental results demonstrate that the proposed method can accurately detect and locate the tire defects.     

一种基于自然图像统计模型的图像合成检测方法

Nowadays, digital images can be easily tampered due to the availability of powerful image processing software. As digital cameras continue to replace their analog counterparts, the importance of authenticating digital images, identifying their sources, and detecting forgeries is increasing. Blind image forensics is used to analyze an image in the complete absence of any digital watermark or signature. Image compositing is the most common form of digital tampering. Assuming that image compositing operations affect the inherent statistics of the image, we propose an image compositing detection method on based on a statistical model for natural image in the wavelet transform domain. The generalized Gaussian model (GGD) is employed to describe the marginal distribution of wavelet coefficients of images, and the parameters of GGD are obtained using maximum-likelihood estimator. The statistical features include GGD parameters, prediction error, mean, variance, skewness, and kurtosis at each wavelet detail subband. Then, these feature vectors are used to discriminate between natural images and composite images using support vector machine (SVM). To evaluate the performance of our proposed method, we carried out tests on the Columbia Uncompressed Image Splicing Detection Dataset and another advanced dataset, and achieved a detection accuracy of 92% and 79%, respectively. The detection performance of our method is better than that of the method using camera response function on the same dataset.     

Color restoration method based on spectral information using normalized cut

This paper proposes a novel method for color restoration that can effectively apply accurate color based on spectral information to a segmented image using the normalized cut technique. Using the proposed method, we can obtain a digital still camera image and spectral information in different environments. Also, it is not necessary to estimate reflectance spectra using a spectral database such as other methods. The synthesized images are accurate and high resolution. The proposed method effectively works in making digital archive contents. Some experimental results are demonstrated in this paper.     

Full-parameter vision navigation based on scene matching for aircrafts

Vision navigation based on scene matching between real-time images and a reference image has many advantages over the commonly used inertial navigation system （INS）, such as no cumulative measurement errors for long-endurance flight. Recent developments in vision navigation are mainly used for partial navigation parameters measurements, such as the position and the relative velocity, which cannot meet the requirements of completely autonomous navigation. We present the concept, principle and method of full-parameter vision navigation （FPVN） based on scene matching. The proposed method can obtain the three-dimensional （3D） position and attitude angles of an aircraft by the scene matching for multiple feature points instead of a single point in existing vision navigations. Thus, FPVN can achieve the geodetic position coordinates and attitude angles of the aircraft and then the velocity vector, attitude angular velocity and acceleration can be derived by the time differentials, which provide a full set of navigation parameters for aircrafts and unmanned aerial vehicles （UAVs）. The method can be combined with the INS and the cumulative errors of the INS can be corrected using the measurements of FPVN rather than satellite navigation systems. The approach provides a completely autonomous and accurate navigation method for long-endurance flight without the help of satellites.     

微小无人直升机自适应视觉伺服

针对摄像机未标定和特征点坐标未知的情况, 本文提出一种新颖的基于图像的无人直升机自适应视觉伺服方法. 控制器是基于反推法设计的, 但是和已有的基于反推法的视觉伺服不同的是, 它利用与深度无关矩阵将图像误差映射到执行器空间, 从而可以避免估计特征点的深度. 这种设计方法可以线性化未知的摄像机参数和特征点坐标, 所以能方便地设计自适应算法来在线估计这些未知参数, 同时为了保证图像误差收敛和避免估计参数收敛至零解而引入了两个势函数. 利用Lyapunov方法证明了基于非线性动力学的控制器的稳定性, 并给出了仿真验证.     

Dynamic Visual Tracking for Manipulators Using an Uncalibrated Fixed Camera

This paper presents a new controller for controlling a number of feature points on a robot manipulator to trace desired trajectories specified on the image plane of a fixed camera. It is assumed that the intrinsic and extrinsic parameters of the camera are not calibrated. A new adaptive algorithm is developed to estimate the unknown parameters online, based on three original ideas. First, we use the pseudoinverse of the depth-independent interaction matrix to map the image errors onto the joint space of the manipulator. By eliminating the depths in the interaction matrix, we can linearly parameterize the closed-loop dynamics of the manipulator. Second, to guarantee the existence of the pseudoinverse, the adaptive algorithm introduces a potential force to drive the estimated parameters away from the values that result in a singular Jacobian matrix. Third, to ensure that the estimated parameters are convergent to their true values up to a scale, we combine the Slotine-Li method with an online algorithm for minimizing the error between the estimated projections and real image coordinates of the feature points. We have proved asymptotic convergence of the image errors to zero by the Lyapunov theory based on the nonlinear robot dynamics. Experiments have been carried out to verify the performance of the proposed controller.     

基于运动视的移动机器人定位方法

提出了一种新颖的基于运动视的定位方法．该方法从机器人接近的目标上选取两个特征点并根据它们的图像坐标确定机器人运动前后相对于目标的位姿．本文还提出了一种能够更准确地找出特征点图像坐标的搜索算法．试验结果表明该方法有较高的定位精度和较强的鲁棒性．     

Uncalibrated visual servoing of robots using a depth-independent interaction matrix

This paper presents a new adaptive controller for image-based dynamic control of a robot manipulator using a fixed camera whose intrinsic and extrinsic parameters are not known. To map the visual signals onto the joints of the robot manipulator, this paper proposes a depth-independent interaction matrix, which differs from the traditional interaction matrix in that it does not depend on the depths of the feature points. Using the depth-independent interaction matrix makes the unknown camera parameters appear linearly in the closed-loop dynamics so that a new algorithm is developed to estimate their values on-line. This adaptive algorithm combines the Slotine-Li method with on-line minimization of the errors between the real and estimated projections of the feature points on the image plane. Based on the nonlinear robot dynamics, we prove asymptotic convergence of the image errors to zero by the Lyapunov theory. Experiments have been conducted to verify the performance of the proposed controller. The results demonstrated good convergence of the image errors.     

多幅图像的自动拼接算法研究

提出一种基于特征点的多幅图像自动拼接算法。根据SIFT或SURF算法在图像的尺度空间中提取特征点,对特征点进行亚像素定位,并赋予主方向。根据特征点邻域信息分布计算得到特征向量后,基于k-d树进行最近邻和次最近邻搜索,利用最近邻特征点距离与次近邻特征点距离之比得到初始匹配点对。使用RANSAC（Random Sample Consensus）算法剔除错误匹配特征点对,同时对图像之间的变换参数进行鲁棒估计,使用多频带融合算法消除拼接痕迹。实验验证了该算法能够完成多幅图像的自动无缝拼接。     

基于特征匹配的单目相机姿态解算

本文采用了一种基于AKAZE特征检测和PnP算法的单目视觉测量方法对相机的相对姿态进行解算,用于快速准确地确定空间中两个目标间的位姿关系.采集合作目标的模板图像,提取附加到合作目标上的4个特征点的像素坐标,利用AKAZE关键点对模板图像和待测图像进行匹配并计算映射矩阵,通过映射矩阵得到4个特征点在待测图像中的像素坐标,然后结合合作目标的尺寸信息求解基于4个共面特征点的PnP问题,解算相机与合作目标的相对位置.实验分析表明该方法计算的实时图像相机位姿与真实结果接近,验证了本文方法的有效性.     

一种新颖的基于两个特征点的室内移动机器人定位方法

提出了一种新颖的基于两个特征点的室内移动机器人定位方法。与已有的几何位姿估计方法或航标匹配方法不同，该方法不需要人工航标，也不需要准确的环境地图，只需一幅由传统的CCD相机拍摄的图像。从机器人接近的目标上选取相对于地面等高的两个点作为两个特征点。基于这两点建立一个目标坐标系。在相机平视且这两个特征点与相机投影中心相对于地面不是恰好等高的条件下，就可以根据这两个特征点在图像中的坐标确定机器人相对于目标坐标系的位置和运动方向。该方法非常灵活，适用范围广，可以大大简化机器人定位问题。试验结果表明这一新的方法不仅简单灵活而且具有很高的定位精度。     

基于立体成像几何特性的动态有限搜索匹配法

提出了一种基于立体成像几何特性的动态有限搜索匹配法。该法根据被测物体与摄像机标定参照物在空间上的相互关系及被测物体形态，确定被测物体特征点视差匹配大致范围。利用立体成像几何特性，采用动态有限搜索法，以求出的视差为参考，根据左(右)图像中特征点与所得匹配对应点的水平视差值，确定在右(左)图像中的搜索方向和搜索范围，采用灰度区域相关计算，确定下一个匹配特征点，并依此类推。由于特征点灰度不稳定，且随视点的不同而有很大差异，为能得到正确匹配，需结合几何相似性约束。通过对一已知三维坐标标准件的计算机仿真实验，证实了该方法的有效性。     

Monocular SLAM with undelayed initialization for an indoor robot

This paper presents a new feature initialization method for monocular EKF SLAM (Extended Kalman Filter Simultaneous Localization and Mapping) which utilizes a 3D measurement model in the camera frame rather than 2D pixel coordinates in the image plane. The key idea is to regard a camera as a range and bearing sensor, of which the range information contains numerous uncertainties. 2D pixel coordinates of measurement are converted to 3D points in the camera frame with an assumed depth. The element of the measurement noise covariance corresponding to the depth of the feature is set to a very high value. And it is shown that the proposed measurement model has very little linearization error, which can be critical for the EKF performance. Furthermore, this paper proposes an EKF SLAM system that combines odometry, a low-cost gyro, and low frame rate (1–2 Hz) monocular vision. Low frame rate is crucial for reducing the price of the processor. This system combination is cost-effective enough to be commercialized for a real vacuum cleaning application. Simulations and experimental results show the efficacy of the proposed method with computational efficiency in indoor environments.     

Image feature detection algorithm based on the spread of Hessian source

Image feature detection can be obtained from many methods including the feature point detection. This paper adopts the image feature point detection method based on second-order characteristics of point and the image feature detection algorithm based on the Hessian matrix to detect more feature points. By combining the gray-scale-based image-matching technology with the feature-based image feature detection technology, we propose a Hessian algorithm to obtain more matching points, which can search for matching more quickly. The proposed algorithm overcomes the traditional matching methods that have Ergodic properties of the search strategy. Experiments demonstrate the speed and accuracy of the proposed algorithm, and we use the correct detected feature points to realize image registration, image fusion and image stitching.