Title: Image recognition method of surface defects of prefabricated concrete members in prefabricated building

Authors: Zhiyuan Li

Addresses: School of Construction Engineering, Chongqing Water Resources and Electric Engineering College, Yongchuan, Chongqing, 402160, China

Abstract: In view of the cracks, holes and other defects on the surface of prefabricated concrete components in prefabricated building, an Image recognition method of surface defects of prefabricated concrete members in prefabricated building is proposed. This method first performs denoising and enhancement processing on the obtained component defect images to improve the clarity of the images. Based on this, the image is segmented, and the defect features of the component surface defect image are extracted based on the segmentation results. Finally, support vector machines are used to classify the extracted features, achieving accurate recognition of surface defects in prefabricated concrete components. The experimental results show that the recognition effect and accuracy of using this method for component defect image recognition are good.

Keywords: prefabricated building; prefabricated concrete members; defect; images recognition.

DOI: 10.1504/IJMPT.2024.136849

International Journal of Materials and Product Technology, 2024 Vol.68 No.1/2, pp.29 - 49

Received: 25 Jun 2023
Accepted: 06 Nov 2023
Published online: 22 Feb 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article