Image Processing of AFM Output Based on Alternating Kalman Filter and Edge Operator

AFM scanning imaging is often affected by various practical factors, resulting in poor image quality, therefore effective image processing is crucial for AFM. Since AFM scanning images are mostly cell images, choosing appropriate edge feature detection operators is particularly critical. In order to better denoise and enhance the algorithm efficiency after AFM scanning, this paper uses the alternating Kalman filter to restore the image. The research shows that the alternating Kalman filter can significantly improve the image restoration effect when row and column are alternately processed. In order to retain more edge details in the restored image, this paper compares the performance of Robert, Log, and Sobel three edge detection operators. Through simulation analysis, the results show that the Log operator performs best in extracting edge features of AFM scanning images. In addition, this paper not only performs comprehensive image processing on cell images, but also applies the method to the processing of nanomaterial images. By validating the scanning images of different samples, further proof is provided that the row-column processing method of alternating Kalman filter and the Log operator are superior in extracting edge features of AFM scanning images. This result shows that the method has wide applicability and excellent performance in image processing of different types of samples.