IDENTIFYING POSSIBLE VOLCANIC ASH LAYERS FROM FLUORESCENT FEATURES IN UV MUDROCK CORE IMAGES USING PYTHON

Humans and computers see images in different ways. Humans combine our sense of sight with how it gets processed by the brain. Whereas, for computers, an image is stacked of numbers. The purpose of this study was to create a code that would only identify possible volcanic ash layers from the UV core image. The RGB scale uses a 3-D matrix values of each color combined to form an image, while the grayscale uses a 2-D matrix with only black and white. The idea was to manipulate the stack of numbers that compose an image in a way that would select only the bright fluorescent layers. There were different types of fluorescence throughput the data, however this study was focused only on the bright fluorescent layers. Five filters were used to process the images: Increase of brightness, grayscale filter, red channel filter, green channel filter and blue channel filter. To conclude, two filter were the best on identifying just the possible volcanic ash layer. The results were based on the purpose of our study and the number of false positives detected by the filters. According to this study the red and the gray channel are the best on identifying the bright fluorescence layers. In addition, all approaches had some successes and some failures, but only the intention of each study would determine how effective the idea is in each case.