To prepare models of automated tasks such as self-driving cars, generation of maps for remote areas or smart surveillance systems are becoming popular nowadays. In this article, the attempts have been made to create Artificial General Intelligence (AGI) to compute distance between objects for automated tasks. In some aspects, machine learning techniques are lacking with rule based systems. Therefore, it is vital to integrate machine learning techniques with rule based systems. There are many tasks which can be done very fast and accurately using rule based algorithms. In some complicated cases such as object detection or face recognition, it may not be a better choice. A real time and rule based approach for computing the distance between objects has been described in this article. This approach is based on algebra, image processing, and optics. It can help to fasten our machine learning algorithms to compute distance between objects. The existing works in this domain are very naïve and the distance of objects was calculated from the camera just by seeing an image. These existing approaches require specific knowledge about the specification of object like height or width or size. However, with the proposed approach, the distance can be computed without any requirement like that. The proposed method successfully computes the distance between objects with commendable performance. The accuracy of proposed method varies in the range of 99.99% to 90.21%. Note to Practitioners: Sophisticated computation of distance is a major requirement in the field of automation tasks. Due to availability of various good quality cameras, one can capture picture of any object from remote location. These pictures can be used for calculating distance of target object from desired location through the proposed method in the paper. The proposed method is calculating distance up to 99.99% accuracy through the use of mathematical and image processing concepts. Through the use of proposed method, the drones can be programmed for capturing picture and calculating its distance for the target object in real time. The devised technique will boost the industry of drone for efficient utilization of its capability. To the best of our knowledge, this is the first attempt of its kind. The computed distance can be used for goods delivery, observing any abnormal activities, prediction of cost in inaccessible hill areas for transportation and traffic management. The limitation of the proposed approach is that the correctness of computed distance depends on the number of cameras. With the increase in the number of cameras, the higher accuracy can be achieved.
Volume 12 | 08-Special Issue
Pages: 968-983
DOI: 10.5373/JARDCS/V12SP8/20202602