Various quantum computing simulation platforms have developed rapidly in the last 3 years. However, few quantum image processing algorithms are simulated in these platforms. In this paper, we design a dual-threshold quantum image segmentation algorithm and simulate it in IBM Q Experience platform through Qiskit extension. The NEQR quantum image representation model is firstly optimized and simulated, which is found that the number of the auxiliary qubits will not increase as the image’s size increases. Then, an efficient quantum comparator to realize the comparison of two numbers is designed. And finally, the high parallelism image segmentation algorithm is proposed and simulated. Suppose the size of an image is \({{2}^{n}}\times {{2}^{n}}\) and the gray-scale scope is [0, \({{2}^{q}}-1\)], the time complexity analysis for the quantum image segmentation algorithm shows that the number of basic quantum gate required is proportional to q and will not increase as image’s size increases. Thus, the proposed quantum segmentation algorithm is highly parallelism and has polynomial time complexity. In addition, the simulation part of this paper will provide reference for other quantum image processing algorithms.

在过去的三年中，各种量子计算仿真平台得到了快速发展。但是，在这些平台上很少模拟量子图像处理算法。在本文中，我们设计了一种双阈值量子图像分割算法，并通过Qiskit扩展在IBM Q Experience平台中对其进行了仿真。首先对NEQR量子图像表示模型进行了优化和仿真，发现随着图像尺寸的增加，辅助量子比特的数量不会增加。然后，设计了一种高效的量子比较器，实现两个数的比较。最后，提出并仿真了高并行度图像分割算法。假设图像的大小是\（{{2} ^ {n}} \乘以{{2} ^ {n}} \），灰度范围是[0，\（{{2} ^ {q }}-1 \）］，对量子图像分割算法的时间复杂度分析表明，所需的基本量子门数量与q成正比，并且不会随着图像尺寸的增加而增加。因此，所提出的量子分割算法是高度并行的并且具有多项式时间复杂度。另外，本文的仿真部分将为其他量子图像处理算法提供参考。