Neuron extraction method and device based on sparse decomposition and depth-of-field estimation

The invention discloses a neuron extraction method and device based on sparse decomposition and depth-of-field estimation. The method comprises the following steps: inputting original image video data; decomposing the original image video data by using a robustness principal component analysis method to obtain a neuron signal, a background and noise; calculating bright and dark channel priors of the neuron signal data, and estimating a transmission rate corresponding to the depth of field of the neuron signal data by using the two priors; removing the transmission rate corresponding to the depth of field from the neuron signal data by taking a pixel as a unit point, and constructing and solving a neuron extraction framework of constraint non-negative matrix factorization; and outputting the extraction information of the neurons, wherein the extraction information comprises the spatial position, the shape size, the action potential and the time sequence change. According to the method, interference information