PCB Test-Pad Detection and Clustering using PySpark and Image Processing

The era of Big Data demands effective and efficient processing and clustering of extensive image data. This paper presents a Spark-based approach to analyze, group, and visualize spatial and color information within image datasets. The method begins with initializing a Spark session, then loading and inspecting the dataset containing the required spatial coordinates (X, Y) and the corresponding RGB color values. We derive statistical insights into the color distributions and perform color-based clustering to identify different color clusters, focusing on isolating greyscale clusters. Then, spatial clustering algorithms are used to derive the patterns in spatial data distribution, facilitating comprehension of color correlations within the spatial settings. Visualizations are created to aid with the understanding and interpretation of clustering results. This methodology presents an effective, scalable solution for analyzing large-scale spatial and colorbased data.