Computer Vision, Machine Learning, and Deep Learning for Wood and Timber Products: A Scopus-Based Bibliometric and Systematic Mapping Review (1983–2026, early access)

Computer vision and image-based analysis have become core technologies for automation, quality control, and optimization in industrial wood and timber processing. Since early defect-detection systems in the 1980s, the field has evolved from rule-based image processing toward machine learning and, more recently, deep learning architectures capable of real-time, high-accuracy performance.

Despite rapid growth—particularly after the adoption of deep learning post-2016—the literature remains fragmented across application domains, sensing modalities, and methodological paradigms. Existing reviews have typically focused on narrow subdomains (e.g., defect detection, species identification, or CT imaging), limited time windows, or single methodological strands. A comprehensive, long-term bibliometric and systematic mapping analysis covering the full industrial scope has been lacking.

This review addresses that gap by quantitatively synthesizing Scopus-indexed research on computer vision, machine learning, and deep learning applied to wood and timber products, spanning four decades of development.