Comparisons of Differential Filtering and Homography Transformation in Modal Parameter Identification from UAV Measurement

This paper proposes a differential filtering method for the identification of modal parameters of bridges from unmanned aerial vehicle (UAV) measurement. The determination of the modal parameters of bridges is a key issue in bridge damage detection. Accelerometers and fixed cameras have disadvantages of deployment difficulty. Hence, the actual displacement of a bridge may be obtained by using the digital image correlation (DIC) technology from the images collected by a UAV. As drone movement introduces false displacement into the collected images, the homography transformation is commonly used to achieve geometric correction of the images and obtain the true displacement of the bridge. The homography transformation is not always applicable as it is based on at least four static reference points on the plane of target points. The proposed differential filtering method does not request any reference points and will greatly accelerate the identification of the modal parameters. The displacement of the points of interest is tracked by the DIC technology, and the obtained time history curves are processed by differential filtering. The filtered signals are input into the modal analysis system, and the basic modal parameters of the bridge model are obtained by the operational modal analysis (OMA) method. In this paper, the power spectral density (PSD) is used to identify the natural frequencies; the mode shapes are determined by the ratio of the PSD transmissibility (PSDT). The identification results of three types of signals are compared: UAV measurement with differential filtering, UAV measurement with homography transformation, and accelerometer-based measurement. It is found that the natural frequencies recognized by these three methods are almost the same. This paper demonstrates the feasibility of UAV-differential filtering method in obtaining the bridge modal parameters; the problems and challenges in UAV measurement are also discussed.

Towards Cross-View Consistency in Semantic Segmentation While Varying View Direction

Several images are taken for the same scene with many view directions. Given a pixel in any one image of them, its correspondences may appear in the other images. However, by using existing semantic segmentation methods, we find that the pixel and its correspondences do not always have the same inferred label as expected. Fortunately, from the knowledge of multiple view geometry, if we keep the position of a camera unchanged, and only vary its orientation, there is a homography transformation to describe the relationship of corresponding pixels in such images. Based on this fact, we propose to generate images which are the same as real images of the scene taken in certain novel view directions for training and evaluation. We also introduce gradient guided deformable convolution to alleviate the inconsistency, by learning dynamic proper receptive field from feature gradients. Furthermore, a novel consistency loss is presented to enforce feature consistency. Compared with previous approaches, the proposed method gets significant improvement in both cross-view consistency and semantic segmentation performance on images with abundant view directions, while keeping comparable or better performance on the existing datasets.

A New Gaze Estimation Method Based on Homography Transformation Derived from Geometric Relationship

In recent years, the gaze estimation system, as a new type of human-computer interaction technology, has received extensive attention. The gaze estimation model is one of the main research contents of the system. The quality of the model will directly affect the accuracy of the entire gaze estimation system. To achieve higher accuracy even with simple devices, this paper proposes an improved mapping equation model based on homography transformation. In the process of experiment, the model mainly uses the “Zhang Zhengyou calibration method” to obtain the internal and external parameters of the camera to correct the distortion of the camera, and uses the LM(Levenberg-Marquardt) algorithm to solve the unknown parameters contained in the mapping equation. After all the parameters of the equation are determined, the gaze point is calculated. Different comparative experiments are designed to verify the experimental accuracy and fitting effect of this mapping equation. The results show that the method can achieve high experimental accuracy, and the basic accuracy is kept within 0.6∘. The overall trend shows that the mapping method based on homography transformation has higher experimental accuracy, better fitting effect and stronger stability.

Deep Features Homography Transformation Fusion Network—A Universal Foreground Segmentation Algorithm for PTZ Cameras and a Comparative Study

The foreground segmentation method is a crucial first step for many video analysis methods such as action recognition and object tracking. In the past five years, convolutional neural network based foreground segmentation methods have made a great breakthrough. However, most of them pay more attention to stationary cameras and have constrained performance on the pan–tilt–zoom (PTZ) cameras. In this paper, an end-to-end deep features homography transformation and fusion network based foreground segmentation method (HTFnetSeg) is proposed for surveillance videos recorded by PTZ cameras. In the kernel of HTFnetSeg, there is the combination of an unsupervised semantic attention homography estimation network (SAHnet) for frames alignment and a spatial transformed deep features fusion network (STDFFnet) for segmentation. The semantic attention mask in SAHnet reinforces the network to focus on background alignment by reducing the noise that comes from the foreground. STDFFnet is designed to reuse the deep features extracted during the semantic attention mask generation step by aligning the features rather than only the frames, with a spatial transformation technique in order to reduce the algorithm complexity. Additionally, a conservative strategy is proposed for the motion map based post-processing step to further reduce the false positives that are brought by semantic noise. The experiments on both CDnet2014 and Lasiesta show that our method outperforms many state-of-the-art methods, quantitively and qualitatively.

Shadow Map Construction using 2D Homography

In computer graphic applications, shadow plays an important role in expressing the reality of an object. Shadow shows the relationship between objects in space. Shadow map is one of the methods that meet the demand simplification in implementation and speed. This method made an alias around the border of shading. In this paper, we proposed a method using a 2D homography transformation. This method reduces the unused area in the shadow maps, so it can help to minimize the alias. With the experiment in VanMieu Tran Bien- a Vietnamese historical place, we compare the method with others.

Image Registration Algorithm Based on Convolutional Neural Network and Local Homography Transformation

In order to overcome the poor robustness of traditional image registration algorithms in illuminating and solving the problem of low accuracy of a learning-based image homography matrix estimation algorithm, an image registration algorithm based on convolutional neural network (CNN) and local homography transformation is proposed. Firstly, to ensure the diversity of samples, a sample and label generation method based on moving direct linear transformation (MDLT) is designed. The generated samples and labels can effectively reflect the local characteristics of images and are suitable for training the CNN model with which multiple pairs of local matching points between two images to be registered can be calculated. Then, the local homography matrices between the two images are estimated by using the MDLT and finally the image registration can be realized. The experimental results show that the proposed image registration algorithm achieves higher accuracy than other commonly used algorithms such as the SIFT, ORB, ECC, and APAP algorithms, as well as another two learning-based algorithms, and it has good robustness for different types of illumination imaging.

An SRTM-Aided Epipolar Resampling Method for Multi-Source High-Resolution Satellite Stereo Observation

Binocular stereo observation with multi-source satellite images used to be challenging and impractical, but is now a valuable research issue with the introduction of powerful deep-learning-based stereo matching approaches. However, epipolar resampling, which is critical for binocular stereo observation, has rarely been studied with multi-source satellite images. The main problem is that, under the multi-source stereo mode, the epipolar-line-direction (ELD) at an image location may vary when computed with different elevations. Thus, a novel SRTM (Shuttle Radar Topography Mission)-aided approach is proposed, where a point is transformed from the original image-space to the epipolar image-space through a global rotation, followed by a block-wise homography transformation. The global rotation transfers the ELDs at the center of the overlapping area to the x-axis, and then block-wise transformation shifts the ELDs of all grid-points to the x-axis and eliminates the y-disparities between the virtual corresponding points. Experiments with both single-source and multi-source stereo images showed that the proposed method is obviously more accurate than the previous methods that do not use SRTM. Moreover, with some of the multi-source image pairs, only the proposed method ensured the y-disparities remained within ±1 pixel.