With the rapid change of urban areas in developing countries, construction areas are constantly appearing in different parts of cities. Those changed areas require timely monitoring to provide up-to-date information f...
详细信息
With the rapid change of urban areas in developing countries, construction areas are constantly appearing in different parts of cities. Those changed areas require timely monitoring to provide up-to-date information for of urban information systems. As a result, it is a challenge to develop an effective change analysis of different objects, especially buildings in cities. This paper presents an object-based framework for analyzing building changes from high-resolution satellite stereo images (HRSSI). The disparity information extracted from stereo images and spectral information including visible vegetation index (VVI) help extract the buildings in a hierarchical approach. Evaluations show the accuracy of higher than 98% and F1-Score of higher than 87% for the building extraction step. Then, each building object is classified into three main categories including "Remained Building", "Removed Building" or "Added Building". Also, the "Remained Building" objects are categorized into four change states including "Only 2d Change", "Only 3d Change", "2d with 3d Change" and "No Change". This is done by utilizing the object-based similarity analysis of the spectral information as well as the similarity analysis of the disparity information using CNN and their integration. Evaluations demonstrate the accuracy of higher than 97% and F1-Score of higher than 90% for this step. (C) 2020 Published by Elsevier Ltd on behalf of COSPAR.
The 6d pose estimation is important for the safe take-off and landing of the aircraft using a single RGB image. due to the large scene and large depth, the exiting pose estimation methods have unstratified performance...
详细信息
The 6d pose estimation is important for the safe take-off and landing of the aircraft using a single RGB image. due to the large scene and large depth, the exiting pose estimation methods have unstratified performance on the accuracy. To achieve precise 6d pose estimation of the aircraft, an end-to-end method using an RGB image is proposed. In the proposed method, the2d and 3d information of the keypoints of the aircraft is used as the intermediate supervision,and 6d pose information of the aircraft in this intermediate information will be explored. Specifically, an off-the-shelf object detector is utilized to detect the Region of the Interest(Ro I) of the aircraft to eliminate backgrounddistractions. The 2d projection and3d spatial information of the pre-designed keypoints of the aircraft is predicted by the keypoint coordinate estimator(Kp Net).The proposed method is trained in an end-to-end fashion. In addition, to deal with the lack of the relateddatasets, this paper builds the Aircraft 6d Pose dataset to train and test, which captures the take-off and landing process of three types of aircraft from 11 views. Compared with the latest Wide-depth-Range method on this dataset, our proposed method improves the average 3ddistance of model points metric(Add) and 5° and 5 m metric by 86.8% and 30.1%, respectively. Furthermore, the proposed method gets 9.30 ms, 61.0% faster than YOLO6d with 23.86 ms.
暂无评论