Research on the Edge Feature Enhancement of Fluvial Reservoirs Based on Image Processing

The identification of channel edge is the key aspect of fine description of fluvial reservoirs. Affected by the factors such as channel overlaying and crossing, thin sand body thickness, low seismic signal-to-noise ratio, and low resolution, the traditional slice interpretation and coherence technology can barely meet the requirements of fine exploration, and the newly developed edge detection based on operator processing still has application misconceptions. In this study, the geometric characteristics of river edges are analyzed as the entry point, and the physical meanings of first derivatives, module values, and second derivatives are clarified. Three-dimensional channel models with different velocity characteristics are established by extracting the coherence attributes of the model and real data; the existing problems of this technique in the fine description of channels are identified. To solve this problem, considering the Sobel operator as an example, the symbolic characteristics of the channel edge after using this technology are illustrated. The coherent enhancement technique for channel edge identification using histogram equalization and the fuzzy set theory is proposed, and good application results are obtained. This method can be used as reference to deepen the understanding of channel edge characteristics and improve the ability to identify fluvial reservoirs.