Digital image processing and integration of data sets have been used to develop exploration models from airborne electromagnetics (EM), magnetics, and very‐low‐frequency electromagnetics (VLF-EM) data collected over an area in northwestern Arizona. The area has potential for the occurrence of uranium‐mineralized breccia pipes. Apparent resistivity and overburden thickness were derived from the EM measurements using half‐space models. Digital image processing techniques applied to the geophysical data sets included: (1) conversion of the data into gridded‐scale images, (2) spatial filtering for noise removal, (3) integration and analysis of the data sets, and (4) modeling using various parameter combinations. The general relationships between the geophysical variables/parameters and their ability to detect metallic deposits were used as guides in selecting digital number ranges that were used as input into various models. One of the best models incorporated apparent resistivity and total‐field magnetics; the results of this model outlined 13 anomalous combinations in the survey area. Field checking confirmed that two of the anomalies were previously known orebodies, and most of the other anomalies corresponded to suspected pipes that were under evaluation by the group that is exploring the property.
Improvement of Convolutional Neural Network Accuracy on Salak Classification Based Quality on Digital Image
