Case Studies

To accurately prepare a bill of quantities and provide a price estimation, the following information needs to be extracted from a floor plan:

• Building type - office building, shopping mall, apartment building
• Floor plan type - electrical, plumbing, etc.
• Number of doors, windows, and rooms
• Total wall length
• Total area

The documents usually contain multiple technical drawings, especially when it comes to multi-story buildings.

Ready-made solutions for optical character recognition (OCR) could not handle the task with enough accuracy or could not handle it at all due to the special characters used in the technical drawings, as well as the graphic nature of the drawings themselves.

The first step in analyzing any PDF file with a floor plan is to detect the location of the floor plan within the page. We have developed a segmentation machine learning model which automatically detects the drawing location. The system also gives users the ability to highlight the floor plan themselves.

The system detects technical drawing type and scale, and automatically generates a table of contents, making it easy to navigate large multi-page documents.

Another important objective we had to achieve is to detect various objects present in the floor plans, like doors, windows, different types of walls, etc., marked by special labels.

As OpenCV algorithms are not well suited for analyzing simple black and white geometric shapes, we have incorporated deep learning to increase the accuracy and reject false positives.

To start with, the user highlights the label that needs to be detected, one for every object group. After that, the processing starts, and the technical drawing is analyzed by the object recognition model. The results are as follows:

• all relevant labels are recognized and counted
• all walls, windows, and doors are recognized and separated into groups according to their properties
• the floor plan is separated into rooms that are counted
• the total area of the floor plan is counted
• the total length of the walls is counted

The user can manually go through the recognition results to fix any mistakes.

Floor plans come with a bill of quantities that contains information about the different labels used in the technical drawing, as well as information about the materials. The PDF tables are not an ideal way to handle large amounts of data since they cannot be edited and the data cannot be sorted or filtered.

There is a number of readymade tools and solutions that can turn a PDF table into an Excel one, though they work poorly with large complex tables that include merged cells and span across multiple PDF pages.

Readymade solutions do not handle merged cells well and often split them incorrectly Often when PDF tables span across multiple pages, their columns don’t line up which causes existing tools to process the data incorrectly If the text goes outside of its cell, readymade solutions split the text into multiple cells

We have developed a subsystem that scans the PDF tables and turns them into Excel tables without changing the original structure of the table and keeping the data integrity.

1. A user uploads a PDF file with a floor plan
2. The system sections of the floor plan and highlights the building type, the floor plan type, and the scale
3. The user corrects any mistakes, highlights the objects that need to be counted, and clicks "Process"
4. The system then recognizes building walls, counts the objects of interest, and provides an Excel file with a bill of quantities and a price estimation.
5. If a bill of quantities is already present in the PDF, the table is converted into an Excel table for easy processing.

A resulting system is a full-fledged tool for working with complex floor plans and accompanying tables. It reduces manual labor and greatly speeds up price estimation. The system is highly flexible and can be adjusted to analyze any PDF document and extract relevant information.