Yimin Deng

Small and dim target detection via lateral inhibition filtering and Artificial Bee colony based selective visual attention.

This paper proposed a novel bionic selective visual attention mechanism to quickly select regions that contain salient objects to reduce calculations. Firstly, lateral inhibition filtering, inspired by the limulus’ ommateum, is applied to filter low-frequency noises. After the filtering operation, we use Artificial Bee Colony (ABC) algorithm based selective visual attention mechanism to obtain the interested object to carry through the following recognition operation. In order to eliminate the camera motion influence, this paper adopted ABC algorithm, a new optimization method inspired by swarm intelligence, to calculate the motion salience map to integrate with conventional visual attention. To prove the feasibility and effectiveness of our method, several experiments were conducted. First the filtering results of lateral inhibition filter were shown to illustrate its noise reducing effect, then we applied the ABC algorithm to obtain the motion features of the image sequence. The ABC algorithm is proved to be more robust and effective through the comparison between ABC algorithm and popular Particle Swarm Optimization (PSO) algorithm. Except for the above results, we also compared the classic visual attention mechanism and our ABC algorithm based visual attention mechanism, and the experimental results of which further verified the effectiveness of our method.

A binocular vision-based UAVs autonomous aerial refueling platform

Unmanned aerial vehicles (UAVs) are highly focused and widely used in various domains, and the capability of autonomous aerial refueling (AAR) becomes increasingly important. Most of the research in this area concerns the verification of the algorithms while the experiments are conducted on the ground. In this work, in order to verify the vision system designed for boom approach AAR, an integrated platform is built and tested. The platform consists of a tanker UAV, a receiver UAV and a ground station. The pictures of the marker on the receiver UAV are captured by the binocular vision system on the tanker UAV and then used for flight control and boom control. Performance and feasibility of the platform are demonstrated by the real out-door flight tests, and the experimental results verified the feasibility and effectiveness of our developed binocular vision-based UAVs AAR.摘要无人机 (Unmanned aerial vehicles, UAVs) 在工程领域受到高度关注和广泛应用, 在这种情况下, 自动空中加油技术 (autonomous aerial refueling, AAR) 的实现则变得越来越重要。 当前, 该领域的研究重点关注算法的实现, 实际验证则多为地面验证。 为了测试硬管加油的视觉系统, 我们设计并开发了一个完整的实验平台。 该平台由一架加油无人机, 一架受油无人机和地面站组成。加油机上搭载的双目视觉系统采集受油机上的标志点图像,将其用于飞行控制和加油杆控制。实际飞行测试的结果验证了该平台的表现和稳定性。文中展示的结果数据证明视觉系统和控制系统可支持硬式自动空中加油技术的实现。

Biological edge detection for UCAV via improved artificial bee colony and visual attention

Purpose – The purpose of this paper is to propose a biological edge detection approach for aircraft such as unmanned combat air vehicle (UCAV), with the objective of making the UCAV recognize targets, especially in complex noisy environment. Design/methodology/approach – The hybrid model of saliency-based visual attention and artificial bee colony (ABC) algorithm is established for edge detection of UCAV. Visual attention can extract the region of interesting objects, and this approach can narrow the searching region for object segmentation, which can reduce the computational complexity. An improved ABC algorithm is applied in edge detection of the salient region. Findings – This work improved ABC algorithm by modifying the search strategy and adding some limits, so that it can be applied to edge detection problem. A hybrid model of saliency-based visual attention and ABC algorithm is developed. Experimental results demonstrated the feasibility and effectiveness of the proposed method: it can guarantee ef...

Biological eagle-eye ¿ Based visual imaging guidance simulation platform for unmanned flying vehicles

An unmanned fying vehicle (UFV) is generally used to search for and track an adversary and even designed to go down with the adversarys vehicle under the extreme conditions of modern wars [1]. The precision of a UFVs electromagnetic sensors generally infuences the accuracy of its guidance. To improve the accuracy of the UFVs guidance system, advanced guidance technologies based on bionic vision have been studied in recent years. Some developments based on bioinspired intelligence [2], [3] have also been investigated for aerial systems. Bionic vision, such as that inspired by the vision of Limulus species (i.e., horseshoe crabs), fruit fies, birds, and humans [4], has become a hot research feld, and some of these new technologies have been applied to missile homing guidance.

Biologically Inspired Model with Feature Selection for Target Recognition Using Biogeography-Based Optimization

To detect salient ground targets precisely and rapidly during aerial reconnaissance, this paper describes a novel object recognition method based on the feature selection of a biologically inspired model and biogeography-based optimization. As a promising approach to object recognition, the biologically inspired model is a hierarchical system of building an increasingly complex and invariant feature representation, which closely follows the process of object recognition in the visual cortex. These scale- and position-tolerant features are constructed by alternating between a template-matching and a maximum-pooling operation. Because of the many patches extracted in the standard biologically inspired model, the random mechanism may extract patches from irrelevant parts of an image and consume a lot of time. In this work, a feature selection method is proposed based on a new population-based evolutionary algorithm called biogeography-based optimization to choose the proper set of patches with high accuracy ...

A binocular vision-based measuring system for UAVs autonomous aerial refueling

In this paper, we present the systematic design and implementation of a binocular vision-based measuring system for autonomous aerial refueling. The hardware configuration of the verification platform is presented, and vision algorithms including feature extraction and pose estimation are employed for estimating the relation between two rotorcrafts. To verify the autonomous aerial refueling of unmanned aerial vehicles, a binocular vision system and an on-board data processing computer are utilized to provide the real-time pose information. Series of experiments are conducted to demonstrate the feasibility and effectiveness of the overall platform.

Close formation flight of swarm unmanned aerial vehicles via metric-distance brain storm optimization

Close formation flight of swarm unmanned aerial vehicles (UAVs) has drawn much attention from scholars due to its significant importance in many aspects. In this paper, we focus on an advanced controller design for swarm UAV close formation based on a novel bio-inspired algorithm, i.e., metric-distance brain storm optimization (MDBSO). The proposed method utilizes the brain storm optimization (BSO) which has been extensively adopted in complicated systems with great performances and modifies its basic operators to formulate the formation flight controller design. The original clustering operator in BSO is replaced by a fresh clustering method based on metric distances, while the individual updating operator utilizes Lévy distribution to extend search steps to fit into the metric searching regions. Then the proposed algorithm is applied to optimize the benchmark controller in swarm UAV close formation to enhance the tracking performances under complicated circumstances. Simulation results demonstrate that our approach is more superior in stable configuration of swarm UAV close formations by comparing with several generic methods.

On-board visual navigation system for unmanned aerial vehicles autonomous aerial refueling

In this paper, an on-board binocular visual navigation system based on unmanned aerial vehicles platform is designed for autonomous aerial refueling. The hardware configuration of the entire platform is introduced. Vision algorithms including target tracking, feature extraction, and pose estimation are employed for measuring the relation between two unmanned aerial vehicles. Two situations are considered under long and short distances to make this procedure flexible. Four pose estimation algorithms are implemented and compared in this research. A series of experiments are conducted to verify the feasibility and effectiveness of aerial refueling on the unmanned aerial vehicle platform.

Drogue pose estimation for unmanned aerial vehicle autonomous aerial refueling system based on infrared vision sensor

Abstract. Autonomous aerial refueling is a significant technology that can significantly extend the endurance of unmanned aerial vehicles. A reliable method that can accurately estimate the position and attitude of the probe relative to the drogue is the key to such a capability. A drogue pose estimation method based on infrared vision sensor is introduced with the general goal of yielding an accurate and reliable drogue state estimate. First, by employing direct least squares ellipse fitting and convex hull in OpenCV, a feature point matching and interference point elimination method is proposed. In addition, considering the conditions that some infrared LEDs are damaged or occluded, a missing point estimation method based on perspective transformation and affine transformation is designed. Finally, an accurate and robust pose estimation algorithm improved by the runner-root algorithm is proposed. The feasibility of the designed visual measurement system is demonstrated by flight test, and the results indicate that our proposed method enables precise and reliable pose estimation of the probe relative to the drogue, even in some poor conditions.

Chaotic mutated bat algorithm optimized edge potential function for target matching

In this paper, we present a novel edge based matching approach to target recognition. To recognize the marker on a rotorcraft, a chaotic mutated bat algorithm optimized edge potential function approach is proposed to accomplish the matching between the sketch image and the scene in real applications. A novel type of attractive contour pattern is acquired using the edge potential function. These edge structures can be conveniently exploited for target matching. Bat algorithm is adopted for the optimization problem of searching the optimal match in the scene, and a chaotic mutated bat algorithm is proposed using the chaotic theory and a mutated operator. Thus, the target matching task is converted to optimizing the average of potential value by the processing of translating, reorienting and scaling the sketch image. Series of experiments are conducted to show that our method is superior to other methods. Our proposed method can achieve the higher fitness value over the standard optimization algorithms.