Search

Despite advancements in computer vision technology for construction site safety, the identification and evaluation of potential safety risks stemming from on-site hazardous objects and their spatial interactions is under-researched. Moreover, the subsequent visualization of risks, essential for effective safety management, remains insufficiently explored. This paper presents a model combining computer vision and the TOPSIS method to automate safety risk quantification and visualization during the construction process. Uniquely, it enhances risk predictions by dynamically monitoring the quantity and distance changes of hazard sources and their risk-related objects in real-time, beyond pre-construction risk identification and predictions. Experimental validation shows the model's effectiveness in risk quantification and visualization, with a high consistency ratio of 95% compared to expert manual assessments. This model lays the groundwork for more precise risk calculation and evaluation, fostering improved safety management decisions. • The use of computer vision technology improves traditional safety risk management. • Proposed model automates safety risk quantification and visualization for site images. • Model assesses impact of spatial position and interactions of hazardous sources on safety. • Risk heatmap generation enhances risk visualization. • Experimental results show highly consistent risk assessments compared to manual assessments. [ABSTRACT FROM AUTHOR]

This paper presents alternative solutions for classifying concrete spall severity based on computer vision approaches. Extreme Gradient Boosting Machine (XGBoost) and Deep Convolutional Neural Network (DCNN) are employed for categorizing image samples into two classes: shallow spall and deep spall. To delineate the properties of a concrete surface subject to spall, texture descriptors including local binary pattern, center symmetric local binary pattern, local ternary pattern, and attractive repulsive center symmetric local binary pattern (ARCS-LBP) are employed as feature extraction methods. In addition, the prediction performance of XGBoost is enhanced by Aquila optimizer metaheuristic. Meanwhile, DCNN is capable of performing image classification directly without the need for texture descriptors. Experimental results with a dataset containing real-world concrete surface images and 20 independent model evaluations point out that the XGBoost optimized by the Aquila metaheuristic and used with ARCS-LBP has achieved an outstanding classification performance with a classification accuracy rate of roughly 99%. • Propose a computer vision method for classifying concrete spall severity. • Extreme Gradient Boosting Machine and deep neural networks are employed. • Use local binary pattern variants as image texture descriptors. • The boosting machine is finetuned by the Aquila optimizer. • The proposed method achieves a high classification accuracy rate = 0.99. [ABSTRACT FROM AUTHOR]

Due to the decentralisation and complexity of knowledge in the architecture, engineering and construction (AEC) industry, the research on knowledge graphs (KGs) is still insufficient, and most of the research focuses on text-based KG structuring or updating. Entity extraction, a sub-task of knowledge extraction, is critical in general KG update approaches. While the mainstream approach for this task generally uses textual data, visual data is more readily available, more accurate and has a shorter update cycle than textual data. Therefore, this paper integrates zero-shot learning techniques with general KGs to present a novel framework called "video2entities" that can extract entities from videos to update the AEC KG. The framework combines the perceptual capabilities of computer vision with the cognitive capabilities of KG to improve the accuracy and timeliness of KG updates. Experimental results demonstrate that the framework can extract "new entities" from architectural decoration videos for AEC KG updates. • A novel video-based framework is proposed for AEC Knowledge Graph (KG) updating. • Two iterative processes are developed for the incremental extraction of entities. • Prior information is strengthened by relationship graphs fusing KG and word embedding. • General KG is updated to AEC domain KG by the application of our framework. [ABSTRACT FROM AUTHOR]

Automatic interpretation of computer aided design drawings into a Building Information Model (BIM) would reduce the time and labor costs involved in the modeling process for construction parties. However, the offset and height of building objects cannot be automatically extracted by current algorithms and products since they mostly focus on floor plan detection whilst such information conventionally appears on elevation drawings. Manual inspection and input are needed, which is rigid, error prone, and costly. The challenge of elevation recognition is attributed to the irregular and intricate shapes of the objects portrayed in elevation views, which make it difficult to fully cluster the primitives composing a building object. Additionally, none of the existing methods includes floor plan detection and elevation detection to enable a comprehensive reconstruction of a 3D BIM model. In this paper, these issues are tackled by resorting to an automatically layer classification method (ALCM) that identifies the content of hidden layers. An ALCM-based elevation recognition method is developed. It recognizes the orientation of elevation views and levels of each floor. Furthermore, it segments openings (windows and doors) in elevation views and outputs their offset and height dimensions. A façade BIM model is generated with all openings placed at the correct offsets. The experiments take 94 different sample drawings to validate the model's performance. The test results demonstrate that nearly all floor levels are detected. And that 88% of the members that are visible in elevation drawings are measured perfectly. A real-world campus building is automatically modelled as a case study. The results imply that ALCM-EDM (Elevation Detection Method) contributes to the automatic conversion process since manual input of elevation data is avoided. Future directions could address on incorporating section views and detailed drawings into the reconstruction. • A novel method to recognize building objects in elevation drawings is developed to automatically reconstruct BIM models. • The method is based on an automatic layer classification method (ALCM) that identifies the content of layers in the CAD system. • The method also detects the orientation of elevation views and floor levels through symbol recognition. • Primitives composing visible openings in elevation views are clustered and height/offset values of openings are measured. • The performance of algorithms is validated by testing 94 sample drawings and a case study of a real-life campus building. [ABSTRACT FROM AUTHOR]