Skip to Main Content
Nowadays, Grayscale Morphological Filter (GMF) is a well-founded non-linear filtering for image processing. Its geometry-oriented nature provides a strong framework for addressing shape characteristics such as size, connectivity, and others, which are not easily accessed by the traditional linear approach. Therefore, grayscale morphological filters have been widely used to enhance and detect the infrared small target, but it is rarely applied to the domain of building target enhancement in dense urban sites. Additionally, its ability to build target enhancements is weak. If the Signal-to-Noise-Ratio (SNR) of the image is low or the simarilty-pattem of the image is high or the image has been contaminated by heavy structured clutters, the traditional grayscale morphological filter may decrease image quality, leading to the loss of target and an increase in false alarms. Therefore, the primary operations of the traditional grayscale morphology, such as erosion, dilatation, opening, closing, and top-hat operations, have drawbacks of creating artificial patterns and distorting or removing significant details. Although they may perform well in some cases, these methods do not really improve the enhancement ability of the grayscale morphological filter.