Abstract

Structural cracks are among the earliest and most reliable visual indicators of deterioration in reinforced-concrete public buildings, yet in Nigeria their detection still relies overwhelmingly on manual visual inspection, a method that is slow, subjective, and frequently unable to prevent catastrophic failure. This study investigated the application of computer vision (CV) and artificial intelligence (AI), particularly convolutional neural network (CNN)-based techniques, for crack detection in public buildings across North Central Nigeria (Federal Capital Territory, Niger, Kwara, Kogi, Benue, Plateau, and Nasarawa states). Three objectives guided the study: to assess the level of awareness of CV/AI-based crack detection technology among builtenvironment professionals; to determine the extent to which CV/AI is applied for crack detection relative to conventional methods; and to identify the challenges limiting its adoption. A descriptive survey design was adopted, and a structured questionnaire was administered to 212 built-environment professionals (architects, civil/structural engineers, builders, and facility managers), of whom 196 (92.5%) returned valid responses. Data were analysed using percentage/weighted-mean analysis for the research questions, and Pearson product-moment correlation and simple linear regression for the hypotheses. Results showed generally low awareness (grand mean = 2.78) and low current application (grand mean = 2.47) of CV/AI crack-detection tools, alongside a high perception of implementation challenges (grand mean = 3.55), principally cost, inadequate infrastructure, and a shortage of skilled personnel. Regression analysis revealed that the extent of CV/AI adoption significantly predicted perceived crack-detection accuracy (R² = 0.421, F(1,194) = 141.15, p < 0.001), while Pearson correlation confirmed significant relationships between awareness and adoption (r = 0.374, p < 0.001) and between challenges and adoption (r = -0.351, p < 0.001). The study concludes that CV/AI technologies offer considerable, largely untapped potential for improving structural safety monitoring of public buildings in North Central Nigeria, and recommends targeted capacity building, infrastructure investment, and policy support to accelerate adoption