Mobile robots lack a driver or a pilot and, thus, should be able to detect obstacles autonomously. This paper reviews various image-based obstacle detection techniques employed by unmanned vehicles such as Unmanned Surface Vehicles (USVs), Unmanned Aerial Vehicles (UAVs), and Micro Aerial Vehicles (MAVs). More than 110 papers from 23 high-impact computer science journals, which were published over the past 20 years, were reviewed. The techniques were divided into monocular and stereo. The former uses a single camera, while the latter makes use of images taken by two synchronised cameras. Monocular obstacle detection methods are discussed in appearance-based, motion-based, depth-based, and expansion-based categories. Monocular obstacle detection approaches have simple, fast, and straightforward computations. Thus, they are more suited for robots like MAVs and compact UAVs, which usually are small and have limited processing power. On the other hand, stereo-based methods use pair(s) of synchronised cameras to generate a real-time 3D map from the surrounding objects to locate the obstacles. Stereo-based approaches have been classified into Inverse Perspective Mapping (IPM)-based and disparity histogram-based methods. Whether aerial or terrestrial, disparity histogram-based methods suffer from common problems: computational complexity, sensitivity to illumination changes, and the need for accurate camera calibration, especially when implemented on small robots. In addition, until recently, both monocular and stereo methods relied on conventional image processing techniques and, thus, did not meet the requirements of real-time applications. Therefore, deep learning networks have been the centre of focus in recent years to develop fast and reliable obstacle detection solutions. However, we observed that despite significant progress, deep learning techniques also face difficulties in complex and unknown environments where objects of varying types and shapes are present. The review suggests that detecting narrow and small, moving obstacles and fast obstacle detection are the most challenging problem to focus on in future studies.

The ability to detect and avoid obstacles is essential for ensuring the safe and efficient navigation of mobile robots. With the increasing demand for intelligent autonomous systems, the need for fast and accurate obstacle detection has become a critical area of research. Existing methods for obstacle detection can be broadly classified into three categories: sensor-based, image-based, and hybrid approaches. Among these, vision-based techniques have gained significant attention due to their effectiveness and versatility. These methods can be further divided into monocular and stereo approaches, each offering distinct advantages. In recent years, stereo vision-based methods have emerged as a promising solution for obstacle detection, as they enable the precise estimation of depth and distance to objects, providing valuable information for real-time navigation. However, despite their accuracy, stereo methods are often criticized for their high computational complexity and processing inefficiency, which can limit their practicality in high-speed robotic applications. This study presents a novel hybrid approach that integrates the advantages of both stereo and monocular methods. By leveraging monocular techniques for object detection and utilizing stereo vision for precise depth estimation, our method enhances efficiency while maintaining high accuracy. Instead of computing depth information for individual pixels, the proposed method operates at the object level, calculating the distance for detected objects rather than processing each pixel separately. In addition, comparative analysis with the method developed by Lamini, Fathi et al. 2024 demonstrates that the proposed approach yields more stable and accurate results, further highlighting its effectiveness and reliability in obstacle detection.