DIGITAL IMAGING IN ENGINEERING AND TECHNICAL DIAGRAMS

DIGITAL IMAGING IN ENGINEERING AND TECHNICAL DIAGRAMS

1. Introduction

Building information modeling is a digital representation of a building’s physical and functional properties and characteristics. This modeling provides every component that goes into a project, which helps the designers, architects, engineers, and construction professionals to create and design more effectively compared to other tools in the market today. This research examines the digital image recognition technology to create a program, which transforms digital images of blueprint diagrams from paper drawings to digital 3D BIM (Building Information Modeling) format. Specifically, it examines the development of pseudocode and a software application that recognizes texts, characters, and objects from 2D image diagrams. Computer-aided design (CAD) software converts the digital image created by software into 3D modeling. Research area of the project includes the pseudocode development of each method as described in flowchart Figure 1 .

2. Research Proposal

Image processing has been successfully used in many areas of civil engineering such as document scanning, satellite imagery to evaluate sites, etc., [1]. Using colors from a picture to detect structural components was introduced by J. Neto in 1990 [2]. Neto et al. [3] described using different images to record and assess the progress of different activities [3]. Digitalized image processing has the potential to be an effective tool when providing information about materials, components, and location of objects that are of interest to professionals in civil engineering. Every component of a building is essentially a BIM (Building Information Modeling) object. This means that the information files for every window, door, or HAVC system of a building are easily accessible. The essence of the building information modeling is that it enables the creation of virtual 3D models that can be explored and manipulated. It makes it easier to understand the relation between spaces, materials, and systems. Digital 3D models allow revisions and the assembly of objects in more accurate details. The objective of this research is to study research available for image processing technology and create a program to convert images to 3D models. As part of the project, this research focuses on the development of machine learning program. The primary aim is to develop a program to learn objects automatically without assistance. Figure 1 illustrates the flow chart process for this research project. The algorithm for machine learning is implemented in each step of the flow chart.

Figure 1: Breakdown of the application flowchart.