Pongsak Khokhuntod. Thread counter based on image processing using a reference signal. Doctoral Degree(Applied Physics). มหาวิทยาลัยนเรศวร. สำนักหอสมุด. : มหาวิทยาลัยนเรศวร, 2024.
Thread counter based on image processing using a reference signal
Abstract:
This study presents a two-part investigation into stripe density analysis based on the Moiré pattern phenomenon. The first part focuses on the theoretical modeling of binary rectangular gratings using logic signals with two discrete levels: "0" and "1." Fourier series are used to describe their frequency components, phase relationships, and duty cycles. The overlap between two gratings is mathematically formulated using expressions that correspond to logical AND and OR operations. Simulation results confirm that the output signals align with the theoretical predictions, showing clear relationships between the structural parameters of the gratings and the characteristics of the resulting Moiré patterns. The second part introduces a practical method for measuring stripe density using image processing techniques. The proposed system applies a moving average filter to isolate low-frequency signals, allowing it to suppress high-frequency noise in both simulated and real fabric images. Grid lines with gradually increasing frequency are generated from left to right and overlaid on the target image to determine where the frequencies match. Experiments conducted on synthetic stripe patterns ranging from 30 to 180 L-1 and on actual fabric images the show that system can measure stripe density with a maximum deviation of less than ±1 L-1 this approach offers a reliable and automated solution suitable for in industrial and scientific applications that require accurate and repeatable stripe density measurements