Procesamiento morfológico de imágenes binarias mediante circuitos cuánticos: una implementación en qiskit

Erosion and dilation are fundamental mathematical morphology operations used to modify the shape of objects in binary images. This study evaluated their implementation through a quantum circuit simulated in Qiskit, using classical processing in OpenCV as the reference. The Baboon image was analyzed at resolutions of 102×102, 128×128, and 256×256 pixels, with 16, 160, 1600, 16000, and 160000 circuit executions. The comparison included visual inspection, calculation of MSE, PSNR, and SSIM, and measurement of simulation times. The circuit reproduced the expected effect of both operations, although the reconstructed images did not fully match those obtained with OpenCV. Dilation showed more stable similarity than erosion. In addition, increasing the number of executions did not improve the metrics monotonically and substantially increased computation time.