CNN-Based Watershed Marker Extraction for Brick Segmentation in Masonry Walls

Abstract

Nowadays there is an increasing need for using artificial intelligence techniques in image-based documentation and survey in archeology, architecture or civil engineering applications. Brick segmentation is an important initial step in the documentation and analysis of masonry wall images. However, due to the heterogeneous material, size, shape and arrangement of the bricks, it is highly challenging to develop a widely adoptable solution for the problem via conventional geometric and radiometry based approaches. In this paper, we propose a new technique which combines the strength of deep learning for brick seed localization, and the Watershed algorithm for accurate instance segmentation. More specifically, we adopt a U-Net-based delineation algorithm for robust marker generation in the Watershed process, which provides as output the accurate contours of the individual bricks, and also separates them from the mortar regions. For training the network and evaluating our results, we created a new test dataset which consist of 162 hand-labeled images of various wall categories. Quantitative evaluation is provided both at instance and at pixel level, and the results are compared to two reference methods proposed for wall delineation, and to a morphology based brick segmentation approach. The experimental results showed the advantages of the proposed U-Net markered Watershed method, providing average F1-scores above 80%.

Wall Stone Extraction Based on Stacked Conditional Gan and Multiscale Image Segmentation

“…However, it is quite burdensome to attach IC tags on each stone and to capture the images individually. As a study similar to our theme, there is one to detect bricks from masonry walls (Ibrahim et al, 2019). This approach combines U-Net based brick seed localization and the Watershed algorithm for accurate instance segmentation of bricks.…”

Automated Dimension Estimation of Steel Pipes Stacked at Construction Sites Using Euclidean Distances Calculated by Overlapping Segmentation

“…The model, trained with a sufficient number of images and repetitions, shows a high recognition rate regardless of the image acquisition environment and the shape of the object and, thus, has been applied to various material management studies. A high recognition rate, regardless of the heterogeneous material, size, shape, and the arrangement of bricks, was reported for a brick-recognition method using a CNN [17]. In addition, a high recognition rate, regardless of the occurrence of rust or overlapping with adjacent rebars, was achieved [18].…”

Image Segmentation of Bricks in Masonry Wall Using a Fusion of Machine Learning Algorithms

“…Although the application of computer vision systems in the construction industry in general is abundant, the literature on the specific application of masonry (brick) wall segmentation is scarce. Some applications employ laser technology (such as terrestrial laser scanners) to sample the environment [23,22,19], while others rely on 24-bit color images [11,16].…”

“…Similarly to our system, Ibrahim et al present a machine learning based segmentation algorithm in [11], however, they only use a single model. Their goal is to segment individual brick instances in a 2D image input, both for modern brick walls as well as ancient archaeological sites.…”

“…Contrary to [23,22,19], our approach is based on color and depth information (from stereo vision) using low-cost imaging hardware. Similarly to [11,16], we use machine learning models, however, we take the pixel-wise classification further and propose an algorithm to combine the segmentation masks of several models.…”