Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Yao Zhao has a total of 22 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Visual Active SLAM Method Considering Measurement and State Uncertainty for Space Exploration

    by Yao Zhao, Zhi Xiong, Jingqi Wang, Lin Zhang, Pascual Campoy

    This paper presents a visual active SLAM method considering measurement and state uncertainty for space exploration in urban search and rescue environments. An uncertainty evaluation method based on the Fisher Information Matrix (FIM) is studied from the perspective of evaluating the localization uncertainty of SLAM systems. With the aid of the Fisher Information Matrix, the Cramér–Rao Lower Bound (CRLB) of the pose uncertainty in the stereo visual SLAM system is derived to describe the boundary of the pose uncertainty. Optimality criteria are introduced to quantitatively evaluate the localization uncertainty. The odometry information selection method and the local bundle adjustment information selection method based on Fisher Information are proposed to find out the measurements with low uncertainty for localization and mapping in the search and rescue process. By adopting the method above, the computing efficiency of the system is improved while the localization accuracy is equivalent to the classical ORB-SLAM2. Moreover, by the quantified uncertainty of local poses and map points, the generalized unary node and generalized unary edge are defined to improve the computational efficiency in computing local state uncertainty. In addition, an active loop closing planner considering local state uncertainty is proposed to make use of uncertainty in assisting the space exploration and decision-making of MAV, which is beneficial to the improvement of MAV localization performance in search and rescue environments. Simulations and field tests in different challenging scenarios are conducted to verify the effectiveness of the proposed method.


  • Covariance Matrix Transformation Method for Absolute/Relative Measurements Fusion of Vision/IMU/GNSS Integration in Parafoil Landing

    by Jingqi Wang, Zhi Xiong, Yao Zhao, Yiming Ding, Yihan Liu, Huanjin Wang, Luqi Yan

    The traditional parafoil system localization methods rely highly on GNSS. However, GNSS cannot provide the control system real-time, high-bandwidth parafoil pose. Additionally, GNSS signals could be attenuated or rejected in complex environments such as mountains or canyons. In order to improve the localization precision of the parafoil system considering GNSS-challenged environments, a multistate constraint Kalman filter-based Vision/IMU/GNSS integration algorithm is explored. In contrast to traditional integration methods defining motion state in a single Cartesian coordinate, our approach jointly defines and constructs the motion state based on multiple coordinates, offering the control system the required high-bandwidth pose input directly. Furthermore, the covariance matrix transformation method is introduced to realize the smooth transition of error state between different coordinates, providing a novel solution for absolute and relative measurement fusion problems. Airsim-based simulation and real-world experiments are carried out to validate the performance of the designed integration algorithm. Experiment results indicate that the proposed algorithm can effectively integrate visual and GNSS measurements, addressing the continuous accurate and effective localization requirement for precise control of autonomous guided parafoil systems in GNSS-challenged environments.


  • Open Access

    Landmark Topology Descriptor-Based Place Recognition and Localization under Large View-Point Changes

    by Guanhong Gao, Zhi Xiong, Yao Zhao, Ling Zhang

    Accurate localization between cameras is a prerequisite for a vision-based heterogeneous robot systems task. The core issue is how to accurately perform place recognition from different view-points. Traditional appearance-based methods have a high probability of failure in place recognition and localization under large view-point changes. In recent years, semantic graph matching-based place recognition methods have been proposed to solve the above problem. However, these methods rely on high-precision semantic segmentation results and have a high time complexity in node extraction or graph matching. In addition, methods only utilize the semantic labels of the landmarks themselves to construct graphs and descriptors, making such approaches fail in some challenging scenarios (e.g., scene repetition). In this paper, we propose a graph-matching method based on a novel landmark topology descriptor, which is robust to view-point changes. According to the experiment on real-world data, our algorithm can run in real-time and is approximately four times and three times faster than state-of-the-art algorithms in the graph extraction and matching phases, respectively. In terms of place recognition performance, our algorithm achieves the best place recognition precision at a recall of 0–70% compared with classic appearance-based algorithms and an advanced graph-based algorithm in the scene of significant view-point changes. In terms of positioning accuracy, compared to the traditional appearance-based DBoW2 and NetVLAD algorithms, our method outperforms by 95%, on average, in terms of the mean translation error and 95% in terms of the mean RMSE. Compared to the state-of-the-art SHM algorithm, our method outperforms by 30%, on average, in terms of the mean translation error and 29% in terms of the mean RMSE. In addition, our method outperforms the current state-of-the-art algorithm, even in challenging scenarios where the benchmark algorithms fail.


  • 3-D Objects Detection and Tracking Using Solid-State LiDAR and RGB Camera

    by Zheng Peng, Zhi Xiong, Yao Zhao, Ling Zhang

    Objects detection and tracking using 3D LiDAR has gained momentum lately while it has not been extensively applied, and the main challenge is that conventional mechanical LiDAR is expensive and it is difficult for single sensor to obtain good tracking performance over a long period of time. In this paper, we propose a multi-sensor fusion 3D objects detection and tracking framework using solid-state LiDAR and RGB camera. We use a low-cost solid-state LiDAR with irregular scan pattern and propose a universal clustering method which determines the searching radius by laser density and range. To improve the overall tracking performance, range and visual information are both utilized to associate 3D objects across different frames. We avoid introducing odometry errors into the system and use visual difference rather than estimated trajectory to distinguish between closely located objects. And we have designed a re-detection process to locate objects that were missed by the clustering algorithm. We evaluated the proposed method on five sequences captured by our ego-vehicle and three additional sequences from open-source datasets. Our results demonstrate that the adaptive searching radius enhances recall and the overall tracking performance is improved by fusing LiDAR and camera.


  • Open Access

    Perception-Aware Planning for Active SLAM in Dynamic Environments

    by Yao Zhao, Zhi Xiong, Shuailin Zhou, Jingqi Wang, Ling Zhang, Pascual Campoy

    This paper presents a perception-aware path planner for active SLAM in dynamic environments using micro-aerial vehicles (MAV). The “Next-Best-View” planner (NBVP planner) is combined with an active loop closing, which is called the Active Loop Closing Planner (ALCP planner). The planner is proposed to avoid both static and dynamic obstacles in unknown environments while reducing the uncertainty of the SLAM system and further improving the accuracy of localization. First, the receding horizon strategy is adopted to find the next waypoint. The cost function that combines the exploration gain and the loop closing gain is designed. The former can reduce the mapping uncertainty, while the latter takes the loop closing possibility into consideration. Second, a key waypoint selection strategy is designed. The selected key waypoints, instead of all waypoints, are treated as potential loop-closing points to make the algorithm more efficient. Moreover, a fuzzy RRT-based dynamic obstacle avoidance algorithm is adopted to realize obstacle avoidance in dynamic environments. Simulations in different challenging scenarios are conducted to verify the effectiveness of the proposed algorithm.


  • KSF-SLAM: A Key Segmentation Frame Based Semantic SLAM in Dynamic Environments

    by Yao Zhao, Zhi Xiong, Shuailin Zhou, Zheng Peng, Pascual Campoy, Ling Zhang


  • Open Access

    A Cooperative Framework Based on Active and Semi-supervised Learning for Sea Ice Classification using EO-1 Hyperion Data

    by Yanling Han, Yao Zhao, Yun Zhang, Jing Wang, Shuhu Yang, Zhonghua Hong, Shouqi Cao

    In the classification of remote-sensing sea ice images, labeled samples are difficult to acquire. To adequately utilize the massive number of unlabeled samples, which contain abundant information, we propose a cooperative framework based on active learning (AL) and semi-supervised learning (SSL) for sea ice image classification. We acquire the most valuable samples using AL and make full use of the abundant information contained in the unlabeled samples using SSL, and then conduct a label consistency verification procedure to further ensure the quality of the pseudo-labeled samples obtained through cooperation between AL and SSL. In the AL part, we adopt a sampling strategy that integrates uncertainty and diversity criteria to acquire the most valuable samples to label. In the SSL part, we utilize the SSL sampling strategy to choose the unlabeled samples with the most information and little redundancy, and use the transductive support vector machine (TSVM) as the classification model. The cooperation between AL and SSL ensures the accuracy of the pseudo-labeled samples through a consistency verification procedure. We conduct comparative experiments using the method proposed and other methods on two hyperspectral images obtained from the Earth Observation Satellite 1 (EO-1). The proposed method achieves the highest classification accuracy for both datasets and can be effectively applied to sea ice classification.