Quantum Clustering Analysis And Quantum Image Recognition

Quantum computing, as a new type of computing model, has become a possible solution to the problem of Moore’s failure. With the in-depth study of quantum computing and machine learning, quantum machine learning also came into being.This paper discusses and studies the quantum clustering algorithm (quantum k-means algorithm) and quantum image recognition in quantum machine learning. The main contents and innovations of this paper are summarized as follows:(1) Based on the classical k-means algorithm, an efficient quantum k-means algorithm based on the principle of distance minimization is proposed. Compared with classical k-means algorithm, the proposed algorithm can bring the exponential speed-up. In order to evaluate the distance between the points to be classified and the centroids of clusters: First to construct the entanglement state of the centroid of cluster and the point to be classified, an auxiliary particle is adjoined. Second, a projective measurement is performed on the auxiliary particle alone. And then the distance between two points based on the result of measurement will be calculated. The goal of the algorithm is to classify the points to be classified to the corresponding clustering according to the minimum distance.(2) The algorithm for comparing the similarity of two quantum images is proposed, and the quantum circuit of the algorithm is given. The proposed algorithm is based on the non-connected images, and the image is represented by quantum states. Next, perform the control swap (c-Swap) operation, and then do quantum measurement, according to the measurement results to determine the similarity of two images.(3) The proposed quantum similarity comparison algorithm is applied to quantum gesture recognition. In the classic field, the process of gesture recognition is more complex. In the quantum field, there is no need to extract the gestures of the color,texture, features and other steps. The gesture can be binarized directly, and then gesture recognition is achieved according to the image similarity algorithm mentioned in (2).