Stereo Vision Based Landmark Detection For Robot Navigation In Unstructured Outdoor Environments

This book is a collection of the papers accepted by the ICIVIS 2021—The International Conference on Image, Vision and Intelligent Systems held on June 15–17, 2021, in Changsha, China. The topics focus but are not limited to image, vision and intelligent systems. Each part can be used as an excellent reference by industry practitioners, university faculties, research fellows and undergraduates as well as graduate students who need to build a knowledge base of the most current advances and state-of-practice in the topics covered by this conference proceedings.

In very few mobile robotic applications stereo vision based navigation and mapping is used because dealing with stereo images is very hard and very time consuming. Despite all the problems, stereo vision still becomes one of the most important resources of knowing the world for a mobile robot because imaging provides much more information than most other sensors. Real robotic applications are very complicated because besides the problems of finding how the robot should behave to complete the task at hand, the problems faced while controlling the robot’s internal parameters bring high computational load. Thus, finding the strategy to be followed in a simulated world and then applying this on real robot for real applications is preferable. In this study, we describe an algorithm for object recognition and cognitive map formation using stereo image data in a 3D virtual world where 3D objects and a robot with active stereo imaging system are simulated. Stereo imaging system is simulated so that the actual human visual system properties are parameterized. Only the stereo images obtained from this world are supplied to the virtual robot. By applying our disparity algorithm, depth map for the current stereo view is extracted. Using the depth information for the current view, a cognitive map of the environment is updated gradually while the virtual agent is exploring the environment. The agent explores its environment in an intelligent way using the current view and environmental map information obtained up to date. Also, during exploration if a new object is observed, the robot turns around it, obtains stereo images from different directions and extracts the model of the object in 3D. Using the available set of possible objects, it recognizes the object.

In robotics, vision is an indispensable tool. It is used for a wide spectrum of applications, such as Simultaneous Localization And Mapping (SLAM), object recognition and communication. Vision is expensive both in terms of hardware (cameras with sufficient image quality, CPU/memory to analyze the images fast enough) and software (need for efficient algorithms). For this reason, robot with advanced vision capabilities are tipically custom-build and expensive. This is particularly true for stereo vision, which has lots of potential. Nevertheless, its use is severely limited by the discussed hardware and software costs. These issues make it difficult to build cheap robots that can achieve stereo vision. A field in which this is one of the main problems is collective robotics, in which the cost of each individual unit is a huge driver in the feasibility of a robot system. Motivated by the above considerations, this work provides the first design of a cheap, affordable robot capable of stereo vision. This work is a part of a bigger project which has additional considerations. In this bigger project, only objects with a set of fixed features are considered interesting for the robot. These objects are single-color objects (although each object can have a different color) and spherical shape. These last considerations lead to the necessity of a color-detection system and three-dimensional position reconstruction based only on spherical targets. The objectives of this work were to create a mobile robotic platform with the discussed capabilities. The platform has been designed and built so that the software and the integration of the hardware modules can be tested. The work described in this document includes both the hardware and the software design and implementation. It presents an innovative way of reconstructing a three-dimensional scene, that is based on color blob-detection. It also provides information about the motion system and the distributed control of the platform. The integration of all the modules has been tested.

In mobile robot applications, it is an important issue for a robot to know where it is. Accurate localization becomes crucial for navigation and map building applications because both route to follow and positions of the objects to be inserted into the map highly depend on the position of the robot in the environment. For localization, the robot uses the measurements that it takes by various devices such as laser rangefinders, sonars, odometry devices and vision. Generally these devices give the distances of the objects in the environment to the robot and proceesing these distance information, the robot finds its location in the environment. In this thesis, two vision-based robot localization algorithms are implemented. The first algorithm uses artificial landmarks as the objects around the robot and by measuring the positions of these landmarks with respect to the camera system, the robot locates itself in the environment. Locations of these landmarks are known. The second algorithm instead of using artificial landmarks, estimates its location by measuring the positions of the objects that naturally exist in the environment. These objects are treated as natural landmarks and locations of these landmarks are not known initially. A three-wheeled robot base on which a stereo camera system is mounted is used as the mobile robot unit. Processing and control tasks of the system is performed by a stationary PC. Experiments are performed on this robot system. The stereo camera system is the measurement device for this robot.

This monograph is devoted to the theory and development of autonomous navigation of mobile robots using computer vision based sensing mechanism. The conventional robot navigation systems, utilizing traditional sensors like ultrasonic, IR, GPS, laser sensors etc., suffer several drawbacks related to either the physical limitations of the sensor or incur high cost. Vision sensing has emerged as a popular alternative where cameras can be used to reduce the overall cost, maintaining high degree of intelligence, flexibility and robustness. This book includes a detailed description of several new approaches for real life vision based autonomous navigation algorithms and SLAM. It presents the concept of how subgoal based goal-driven navigation can be carried out using vision sensing. The development concept of vision based robots for path/line tracking using fuzzy logic is presented, as well as how a low-cost robot can be indigenously developed in the laboratory with microcontroller based sensor systems. The book describes successful implementation of integration of low-cost, external peripherals, with off-the-shelf procured robots. An important highlight of the book is that it presents a detailed, step-by-step sample demonstration of how vision-based navigation modules can be actually implemented in real life, under 32-bit Windows environment. The book also discusses the concept of implementing vision based SLAM employing a two camera based system.

Localization and mapping are the essence of successful navigation in mobile platform technology. Localization is a fundamental task in order to achieve high levels of autonomy in robot navigation and robustness in vehicle positioning. Robot localization and mapping is commonly related to cartography, combining science, technique and computation to build a trajectory map that reality can be modelled in ways that communicate spatial information effectively. This book describes comprehensive introduction, theories and applications related to localization, positioning and map building in mobile robot and autonomous vehicle platforms. It is organized in twenty seven chapters. Each chapter is rich with different degrees of details and approaches, supported by unique and actual resources that make it possible for readers to explore and learn the up to date knowledge in robot navigation technology. Understanding the theory and principles described in this book requires a multidisciplinary background of robotics, nonlinear system, sensor network, network engineering, computer science, physics, etc.