Inspection of edge detection using spatial frequency filtering

This study proposes a defect detection system based on spatial frequency filtering to measure edge chipping in two types of tools: straight-edged punching blades and closed-curve punching dies, both made of ultrafine grain cemented carbide. Although different optical configurations were applied to accommodate the geometry of each tool, both systems are based on the same measurement principle using high-pass spatial frequency filtering, and they share a unified image processing algorithm for quantifying edge chipping and suppressing the influence of dust and other noise. A spatial filter is placed at the back focal plane of the Fourier transform lens to block the 0-order ray, forming a clear light-dark-light fringe pattern. Combined with the proposed image processing algorithm, this enables accurate extraction of edge positions and robust evaluation of micro-scale defects. Experiments demonstrated that the developed system could detect edge chipping larger than 1 μm in both types of tools. This research presents a practical and adaptable solution for non-contact defect detection and contributes to advanced quality assurance in industrial blade manufacturing. • High-precision edge chipping detection using spatial frequency filtering technology. • Detects defects larger than 3 microns, unaffected by dust or debris near blade edges. • Achieves fast, automated full-circumference defect detection in complex-shaped tools. • System enables high-speed, submicron-scale edge inspection for various blades.