In this paper, we propose an algorithm, named hashingbased non-maximum suppression (HNMS) to efficiently suppress the nonmaximum boxes for object detection. Non-maximum suppression (NMS) is an essential component to suppress the boxes at closely located locations with similar shapes. The time cost tends to be huge when the number of boxes becomes large, especially for crowded scenes. The basic idea of HNMS is to firstly map each box to a discrete code (hash cell) and then remove the boxes with lower confidences if they are in the same cell. Considering the intersection-over-union (IoU) as the metric, we propose a simple yet effective hashing algorithm, named IoUHash, which guarantees that the boxes within the same cell are close enough by a lower IoU bound. For two-stage detectors, we replace NMS in region proposal network with HNMS, and observe significant speed-up with comparable accuracy. For one-stage detectors, HNMS is used as a prefilter to speed up the suppression with a large margin. Extensive experiments are conducted on CARPK, SKU-110K, CrowdHuman datasets to demonstrate the efficiency and effectiveness of HNMS. Code is released at https://github.com/microsoft/hnms.git.

Epipolarity is the foundation of epipolar resampling, which is used to eliminate the vertical disparity between stereo pairs in stereo matching. To analyze the epipolarity of pushbroom satellite images, this paper compares three methods based on the rational function model (RFM): the projection trajectory method (PTM), the piecewise projection trajectory method (PPTM) and our extended projection trajectory method (EPTM). To evaluate the quality of epipolar curves, we defined the deviation coefficient as a metric to evaluate the bending degree of epipolar curves. We also defined the maximum deviation coefficient of an image that can be used to determinate the tile size in multiview satellite image 3D reconstruction based on image dividing. Comparison experiments have been carried out with pushbroom satellite images using these three methods. Experimental results show that our EPTM is more convenient and practical. It only needs the forward form of the RFM to analyze the epipolarity and can be used in the epipolarity analysis of a single image. By projecting straight lines in the ground space into the image space, the EPTM can be used to perform comprehensive epipolarity analysis for pushbroom satellite images. In addition, the EPTM can be used to calculate the maximum deviation coefficient of an image that the PTM and the PPTM cannot calculate, which is important in 3D reconstruction using multiview satellite images. INDEX TERMS Epipolarity; stereo matching; pushbroom satellite image; RFM; projection trajectory method; deviation coefficient.

Aiming at the problems of insufficient network fusion and low detection efficiency in current object recognition using RGB-D images, a recognition algorithm based on the medium-level layer-by-layer fusion of dual-channel networks is proposed. First of all, the RGB and Depth networks are trained with ten labelled RGB-D indoor objects respectively, and then determine the fusion coefficients according to the identify accuracy of two types networks. Finally, two kinds of features are merged in convolutional layers step by step to obtain the final weights. By testing on the challenging NYU Depth v2 dataset, we found that the recognition accuracy of our method is 92.85%, and average detection time is 61.03ms per image. Through comparison experiments, we got the conclusion that average accuracy of the RGB-D layer-by-layer fusion network is 5.22% higher than that of the RGB network.

No abstract

Abstract. Satellite imaging direction angles, including the azimuth angle and the incidence angle, are the basic information used for satellite camera network structure analysis. They play an important role in 3D reconstruction using satellite images. In this paper, a satellite imaging direction angle estimation method based on rational polynomial coefficients is proposed for use when the satellite imaging direction angles are not available. Using rational polynomial coefficients, a vertical line on the ground is projected into the image plane, and the satellite imaging direction angles are estimated by analyzing the projection. Satellite images acquired by SPOT6, SOPT7 and Pleiades with different satellite imaging direction angles were used to test the feasibility of the proposed method. The experimental results were analyzed in detail combined with the method and the data. The experimental results show that the azimuth angle estimation error is less than 1.30 degrees, and the incidence angle estimation error is less than 0.83 degrees. This level of accuracy is sufficient for satellite camera network structure analysis.