An adaptive threshold-based quantum image segmentation algorithm and its simulation

Efficient and accurate image segmentation algorithm is critical to image processing. In this paper, we design a quantum image segmentation algorithm utilizing an adaptive threshold based on a moving average method, and we simulate it on the IBM Quantum Experience (IBM Q) platform through the Qiskit extension. In the proposed method, an image is first divided into many 2 OE 2 regions, and each region's average value is considered the region's threshold value. In order to fully exploit quantum parallelism, we encode the core image (image to be segmented) and the three auxiliary images into one quantum superposition state sharing the same position qubits. The analysis results highlight that the proposed quantum image segmentation algorithm provides exponential speedup over the existing implementations, and the number of auxiliary qubits is reduced from exponential of q to polynomial. In addition, this paper presents an appealing example of simulating complex quantum image processing algorithms in quantum simulators.

Automated summary

This paper presents a quantum image segmentation algorithm that utilizes quantum parallelism to provide exponential speedup compared to existing implementations, while reducing the number of auxiliary qubits.