ID 原文 译文
10454 实验仿真结果证明,该算法在运动目标存在缩放、旋转等形变时,仍可以准确跟踪目标。 Experimental simulation results show that the algorithm in the moving targets exist the deformation such as scaling, rotation, still can accurately track the target.
10455 场景语义分类是图像理解领域中一个重要的研究方向,涉及到信号处理、模式识别、计算机视觉和认知科学等多学科交叉。 Scene the semantic classification is an important research in the area of image understanding, involves signal processing, pattern recognition, computer vision and cognitive science and other multi-disciplinary cross.
10456 场景分类任务中,图像内容描述和分类判决是两大关键问题。 ‭Scene classification task, image content description and classification judgment are two key problems. ‭
10457 图像内容描述力图得到关于场景图像最具判别意义的表示,而分类判决则对训练样本集的图像内容描述学习、训练,并建模得到某类场景图像区别于其他场景类图像的计算模型。 Image content description is trying to get images of the most discriminant significance said that while the classification decision image content description of the training sample set of learning, training, and get some sort of scene modeling image is different from other types of calculation model of the image scene.
10458 目前,很多场景分类方法针对图像内容描述和图像分类进行了深入的研究,对室外人造场景、室外自然场景和室内场景图像进行分类,取得了较好的分类效果。 ‭At present, a lot of scene classification method for image content description and image classification conducted in-depth research, for the artificial outdoor scene, outdoor natural scene image classification and indoor scene, a better classification results have been achieved.
10459 然而,场景图像自身内容上的变化和差异,既会造成同一场景类内对象的差异性,同时也造成不同场景类之间图像的视觉相似性,特别是对于不同的室内场景类。 ‭Images, however, the change of its content and differences, both can cause the difference of the objects within the same scene class, but also lead to different scene image visual similarity between classes, particularly for indoor scene in different classes.
10460 因此,场景语义分类任务十分困难,是计算机视觉和认知心理学领域中一个颇具挑战性的难题。 ‭Scene semantic classification task is very difficult, therefore, in the field of computer vision and cognitive psychology is a challenging problem.
10461 室外图像场景分类研究相对成熟,而室内图像场景分类研究却进展缓慢。 ‭Outdoor image scene classification research relatively mature, while indoor image scene classification research progress has been slow.
10462 本文综述了图像场景语义分类的研究进展,并分析了场景分类算法的性能,指出场景语义分类研究中存在的问题。 ‭This paper reviews the research progress of scene image semantic classification, and analyze the performance of scene classification algorithm, points out existing problems in the research of semantic classification scene. ‭
10463 针对具有非线性不确定性、不可测动态不确定性和参数不确定性的非线性系统,结合反演提出了自抗扰控制器。 With nonlinear uncertainties and unpredictable dynamic uncertainties and parameter uncertainty of nonlinear system, combined with inversion (adrc) is proposed.