Towards a quantitative analysis of magnetic force microscopy data matrices

Abstract Fast and efficient software tools previously developed in image processing were adapted to the analysis of raw datasets consisting of multiple stacks of images taken on a sample interacting with a measuring instrument and submitted to the effect of an external parameter. Magnetic force microscopy (MFM), a follow-up of atomic force microscopy (AFM), was selected as a first testbed example. In MFM, a specifically developed ferromagnetic scanning tip probes the stray magnetic field generated from a ferromagnetic specimen. Raw scanning probe images taken on soft patterned magnetic materials and continuous thin films were used, together with synthetic patterns exploited to assess the absolute performance ability of the proposed texture analysis tools. In this case, the parameter affecting the sample-instrument interaction is the applied magnetic field. The application discussed here is just one among the many possible, including, e.g., real-time microscopy images (both optical and electronic) taken during heat treatments, phase transformations and so on. Basically any image exhibiting a texture with a characteristic spatial or angular dependence could be processed by the proposed method. Standard imaging tools such as texture mapping and novel data representation schemes such as texture analysis, feature extraction and classification are discussed. A magnetic texture stability diagram will be presented as an original output of the entropic analysis on MFM datasets.