Image Interpretations Based on Quantum Resonance Concept

The study integrates the resonance principle of quantum physics and random walks concept to analyze the image multi-understandings. The quantum model of the photo and bounded electron interaction is recruited to simulate the machine image understanding problems. The pixel or the unit cell of the image will be resonance when the pixel or the unit cell satisfied the given quantum condition. Besides, the random walk concept is introduced to choose which area of the image will be proceeded the resonance process. Basically, the different sampling patterns will introduce the multiple comprehensions. The properties of resonance map; e.g. spatial isotropic, spatial homogenous, geometric condition, gray level distribution; could be taken as the significant features to identify the resonance status. In simulation, the combined characters C, O, Chinese character "jade", and "king" are employed. The results reveal the different sampling patterns can lead multi resonance outputs.