| Spectrogram analysis is an efficient way to analyze the signal changing and locate where the change occurs. Normally, the Fourier Transform is performed on a short interval of time or spatial distance to characterize the non-stationary dataset. By recognizing the unique patterns which are produced by the audio spectrograms, it has been successfully used in audio signal analysis such as recognizing human voices and learning an animal language. The significant feature of an audio dataset is that the audio wave changes periodically. The Fourier Transform is a powerful tool to recognize the components of the audio wave with different frequencies. Based on a high-density sampled dataset, an audio spectrogram is created by performing a Fast Fourier Transform on each short interval of the input signal. Similarly, spatial spectrogram analysis can be used to analyze a spatial dataset if the dataset has the same feature as the audio dataset.; Because of the technologies used to collect the spatial datasets, the spatial dataset is sampled sparsely and irregularly in both Easting and Northing spatial coordinates. Hence, we cannot produce spatial spectrogram by performing the Fourier Transform on each spatial interval directly. To overcome this shortcoming, we organize the data for one dimension, e.g., Easting coordinate. (Abstract shortened by UMI.)... |
