A front-end technique for visual gold detection and localization – Towards automation of the gold panning process.

Abstract

To do away with the time consuming, error prone and expert dependent gold detection and localization in the current gold panning procedure, automatic image-based techniques are considered in this paper. Identifying and locating gold particles during the gold panning process is a fundamental process in gold extraction. In the automation of the gold panning process and robotic handling it is important to identify and locate the gold particles in the images captured by the image sensor. Image segmentation is a vital step in image simplification, image understanding and object detection. Image segmentation is the process of identifying and extracting homogeneous regions (segments) in an image satisfying a homogeneity criterion based on features formulated from spectral components of the image. Three image thresholding techniques were tested and evaluated on sample gold panning images. Color image thresholding in the CIELAB color space performed better in detecting and locating gold particles in an image. The proposed method will serve as a front-end technique for an automated gold panning system as it will automate the visual feature identification of gold particles and aid in the control of the handling system of the gold particles during the panning process.