Search

目前同时定位与地图构建技术多假设外部环境是静态的，忽略了动态目标对SLAM系统的影响，这一假设很大程度上影响了无人系统自主导航的定位精度与鲁棒性。针对这一问题，提出了一种结合目标检测语义信息与多视图极线几何约束的动态SLAM系统，根据运动概率来判别与剔除系统中的动态特征点，并在TUM数据集及实机平台上进行了算法性能测试。实验结果表明在高动态环境下，该系统相较于ORB-SLAM2在绝对轨迹误差上至少可减小94%以上，相对位移轨迹和旋转角度误差分别至少降低了41%和40%，表明该算法能够有效剔除动态特征点，提高了动态环境下视觉SLAM系统的定位精度和鲁棒性。

Accurately segmenting pancreas or pancreatic tumor from limited computed tomography (CT) scans plays an essential role in making a precise diagnosis and planning the surgical procedure for clinicians. Although deep convolutional neural networks have greatly advanced in automatic organ segmentation, there are still many challenges in solving the pancreas segmentation problem. To solve the segmentation problem of pancreas and pancreatic tumors with small areas and complex background, a new coarse-to-fine segmentation framework based on transform function and active localization offset was considered. Specifically, CT slices were input to the coarse segmentation network, and then the visual cues obtained by the coarse segmentation network were converted into spatial weights through a conversion function. At the same time, during the iterative execution of the coarse segmentation network, the ALOT module dynamically adjusted the localization of the target object, and obtained the cropped area with richer semantic information around the target to input to the fine segmentation network. Finally, the finer segmentation results were output by the fine segmentation network. Experiments in the NIH pancreas segmentation dataset and the MSD pancreatic tumor segmentation dataset shows that the framework achieves 85.15% (DSC) and 63.36% (DSC) on the NIH and MSD datasets, respectively, which are 0.25%, 0.5% higher than the previous best method, and the calculation time and calculation amounts only need about 1/3 of the best method. Code will be available on the author′s github personal homepage. [ABSTRACT FROM AUTHOR]

Copyright of Shipin Kexue/ Food Science is the property of Food Science Editorial Department and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Transactions of the Chinese Society of Agricultural Engineering is the property of Chinese Society of Agricultural Engineering and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

When the information conditions of the crashed aircraft are default, locating the aircraft search area is extremely important to carry out the follow-up rescue work. In order to improve the positioning accuracy, through the relevant theory of the decision field, the pilot's emergency behavior decision-making during the forced landing process was analyzed, and an aircraft search range setting method that coupled the pilot's emergency behavior decision-making and flight dynamics was proposed. On this basis, the flight environment was established according to the meteorological and topographical conditions in the high plateau area. Taking the B737-700 model as an example, taking into account the differences in the pilot's decision-making results, the forced landing flight process of the aircraft with some default information conditions was simulated. The scope and priority of the search area were determined accordingly. The simulation results show that in the case of unstable factors or default information, the method of coupling the pilot's emergency behavior decision-making with the dynamic model is reliable. [ABSTRACT FROM AUTHOR]