A Bio-inspired Collision Avoidance Model Based on Spatial Information Derived from Motion Detectors Leads to Common Routes.
A Bio-inspired Collision Avoidance Model Based on Spatial Information Derived from Motion Detectors Leads to Common Routes.
复制标题
DOI：
10.1371/journal.pcbi.1004339
复制
发表时间：
2015-11
影响因子：
4.3
通讯作者：
Egelhaaf M
中科院分区：
文献类型：
--
作者：
Bertrand OJ;Lindemann JP;Egelhaaf M
避免碰撞是任何移动的代理的最基本的需求之一，无论是生物的还是技术的，当搜索周围或瞄准一个目标时。我们提出了一个模型的碰撞避免的灵感来自昆虫的行为实验和翻译过程中经历的球形眼睛的光流特性，并测试该模型与目标驱动的行为的相互作用。昆虫，如苍蝇和蜜蜂，通过行为，即通过采用飞行和凝视控制的扫视策略，主动分离旋转和平移光流分量。在平移期间，即在扫视间阶段期间经历的光流包含关于环境的深度结构的信息，但是该信息与关于自运动的信息纠缠在一起。在这里，我们提出了一个简单的模型来提取的深度结构，通过使用本地属性的球面眼睛的平移光流。在此基础上，计算确保碰撞避免的代理的运动方向。飞行昆虫被认为是通过相关型基本运动探测器来测量光流的。它们的反应除了取决于速度外，还取决于物体的质地和对比度，因此不能真实地测量物体的速度。因此，我们最初使用几何确定的光流作为碰撞避免算法的输入，以表明从光流推断的深度信息足以说明闭环条件下的碰撞避免。然后，在其输入中使用生物启发的相关型基本运动检测器对碰撞避免算法进行测试。即使这样，该算法成功地避免了碰撞，此外，复制了昆虫的避碰行为的特征。最后，将避碰算法与目标方向相结合，并在复杂环境中进行了测试。然后，模拟的代理显示出目标导向的行为，让人想起昆虫导航行为的组成部分。我们周围的机器人数量正在不断增加。它们被用来拯救人类，检查危险地形或清洁我们的家园。在过去的几十年里，它们变得更加自主，更安全，建造成本更低。每个自主机器人都需要在复杂的环境中导航，而不会与沿着的障碍物发生碰撞。目前，他们大多使用主动传感器来解决这个问题，这导致相对较高的能量成本。然而，飞行昆虫能够主要依靠视觉来解决这个任务。任何生物和技术的媒介物在穿过环境时，都会在视网膜上经历环境的明显运动。表观运动包含了自运动和智能体到环境中物体的距离的纠缠信息。后者对于避免碰撞至关重要。从几何视运动中提取物体的相对距离是一项相对简单的任务。然而，试图用生物运动检测器（即在动物王国中发现的运动检测器）来实现这一点是棘手的，因为它们不提供明确的速度信息，但也受到环境的纹理属性的影响。受昆虫能力的启发，我们开发了一种简约算法，仅基于基本运动检测器来避免在具有挑战性的环境中发生碰撞。我们将我们的算法与目标方向相结合，然后在杂乱的环境中进行测试。从这个算法产生的轨迹显示出有趣的目标导向的行为，如形成少量的路线，也观察到导航昆虫。
Avoiding collisions is one of the most basic needs of any mobile agent, both biological and technical, when searching around or aiming toward a goal. We propose a model of collision avoidance inspired by behavioral experiments on insects and by properties of optic flow on a spherical eye experienced during translation, and test the interaction of this model with goal-driven behavior. Insects, such as flies and bees, actively separate the rotational and translational optic flow components via behavior, i.e. by employing a saccadic strategy of flight and gaze control. Optic flow experienced during translation, i.e. during intersaccadic phases, contains information on the depth-structure of the environment, but this information is entangled with that on self-motion. Here, we propose a simple model to extract the depth structure from translational optic flow by using local properties of a spherical eye. On this basis, a motion direction of the agent is computed that ensures collision avoidance. Flying insects are thought to measure optic flow by correlation-type elementary motion detectors. Their responses depend, in addition to velocity, on the texture and contrast of objects and, thus, do not measure the velocity of objects veridically. Therefore, we initially used geometrically determined optic flow as input to a collision avoidance algorithm to show that depth information inferred from optic flow is sufficient to account for collision avoidance under closed-loop conditions. Then, the collision avoidance algorithm was tested with bio-inspired correlation-type elementary motion detectors in its input. Even then, the algorithm led successfully to collision avoidance and, in addition, replicated the characteristics of collision avoidance behavior of insects. Finally, the collision avoidance algorithm was combined with a goal direction and tested in cluttered environments. The simulated agent then showed goal-directed behavior reminiscent of components of the navigation behavior of insects. The number of robots in our surroundings is increasing continually. They are used to rescue humans, inspect hazardous terrain or clean our homes. Over the past few decades, they have become more autonomous, safer and cheaper to build. Every autonomous robot needs to navigate in sometimes complex environments without colliding with obstacles along its route. Nowadays, they mostly use active sensors, which induce relatively high energetic costs, to solve this task. Flying insects, however, are able to solve this task by mainly relying on vision. Any agent, both biological and technical, experiences an apparent motion of the environment on the retina, when moving through the environment. The apparent motion contains entangled information of self-motion and of the distance of the agent to objects in the environment. The later is essential for collision avoidance. Extracting the relative distance to objects from geometrical apparent motion is a relatively simple task. However, trying to accomplish this with biological movement detectors, i.e. movement detectors found in the animal kingdom, is tricky, because they do not provide unambiguous velocity information, but are much affected also by the textural properties of the environment. Inspired by the abilities of insects, we developed a parsimonious algorithm to avoid collisions in challenging environments solely based on elementary motion detectors. We coupled our algorithm to a goal direction and then tested it in cluttered environments. The trajectories resulting from this algorithm show interesting goal-directed behavior, such as the formation of a small number of routes, also observed in navigating insects.