Automatic Salt Segmentation with UNET in Python using Deep Learning

Salt – affected lands are not used for agriculture. The presence of salt within the soil solution reduces the flexibility of the plant to absorb water, and this ends up in reductions within the rate of growth. So, the salt segmentation is being done to find the land containing salt. Segmentation of salt deposits beneath the Earth surface, where the seismic image which are of a particular pixel, that pixel is either classified as salt or sediment. The goal of this salt segmentation project is to segment region that contains salt. UNET architecture is used to achieve the segmentation results. From the input image, we select the important part. The idea is to retain only the important features from the given region image. UNET architecture contains two paths. The first path is the contraction path which is used to capture the context in the image. The second path is the symmetric expanding path which is used to enable precise localization using transposed convolutions. By using this architecture, we can find salt segmentation, from the seismic will get the salt region, then the value of the total salt region is been shown.