The sky region in an image provides horizontal and background information for autonomous ground robots and is important for vision-based autonomous ground robot navigation. This paper proposes a sky region detection algorithm within a single image based on gradient information and energy function optimization. Unlike most existing methods, the proposed algorithm is applicable to both colour and greyscale images. Firstly, the gradient information of the image is obtained. Then, the optimal segmentation threshold in the gradient domain is calculated according to the energy function optimization and the preliminary sky region is estimated. Finally, a post-processing method is applied in order to refine the preliminary sky region detection result when no sky region appears in the image or when objects extrude from the ground. Experimental results have proven that the detection accuracy is greater than 95% in our test set with 1,000 images, while the processing time is about 150ms for an image with a resolution of 640×480 on a modern laptop using only a single core.

No abstract

No abstract

Deep learning approach has been used extensively in image analysis tasks. However, implementing the methods in 3D data is a bit complex because most of the previously designed deep learning architectures used 1D or 2D as input. In this work, the performance of deep learning methods on different 3D data representations has been reviewed. Based on the categorization of the different 3D data representations proposed in this paper, the importance of choosing a suitable 3D data representation which depends on simplicity, usability, and efficiency has been highlighted. Furthermore, the origin and contents of the major 3D datasets were discussed in detail. Due to growing interest in 3D object retrieval and classification tasks, the performance of different 3D object retrieval and classification on ModelNet40 dataset were compared. According to the findings in this work, multi views methods surpass voxel-based methods and with increased layers and enough data augmentation the performance can still be increased. Therefore, it can be concluded that deep learning together with a suitable 3D data representation gives an effective approach for improving the performance of 3D shape analysis. Finally, some possible directions for future researches were suggested.

Abstract-In this paper, we propose a novel H-infinity filter based particle filter (H∞PF), which incorporates the H-infinity filter (H∞F) algorithm into the particle filter (PF). The basic idea of the H∞PF is that new particles are sampled by the H∞F algorithm. Since the H∞F algorithm can fully take into account the current measurements, when the new algorithm calculates the proposed probability density distribution, the sampling particles can take advantage of the system current measurements to predict the system state. The particles distribution we obtained approaches nearer to the state posterior probability distribution and the H∞PF alleviates the sample degeneracy problem which is common in the PF, especially when the maneuvers of the target tracking are large. Furthermore, the H∞F algorithm can adjust gain imbalance factor by adjusting disturbance attenuation factor, from that the new algorithm can get the compromise between the accuracy and robustness and we can obtain satisfied accuracy and robustness. Some simulations and experimental results show that the proposed particle filter performed better than the PF and the Kalman particle filter (KPF) in tracking maneuvering target.