Reverse Engineering in DTs Using Object Detection

Abstract Object detection has become a fairly widespread method in recent years for computer processing that uses artificial intelligence to detect certain sought-after objects in digital images or videos. Object detection and recognition techniques have also experienced continuous development as their implementa-tion, use, and benefits have become extremely evident through the speed with which answers are found in the images submitted to attention. Thus, applications that use object recognition have become very diverse: recognition of people, au-tonomous cars, construction of robots, voice recognition, automatic translation of texts and spoken messages, robotic surgery and patient-centered medical care, recognition of components in industry, recognition of manufacturing defects, recognition of crops, etc. This paper is a further step of previous research that proposed innovative methods for object recognition using advanced CAD tools (CAX), when the custom data-base does not contain images of objects, but CAD models of them. This step studies the use of object recognition in Reverse Engineering methods, i.e., updat-ing a CAD model based on object detection.