A De-noising Method of Difference Pattern based on Combined Filtering Algorithm

If there are abnormal points in the difference pattern, the calculation results of electrical performance indicators will be affected and deviate from the true value. In order to solve the problem of large errors in the processing results of abnormal electrical performance test data, a combined filtering algorithm based on mean filtering and wavelet transform is proposed in this paper to effectively deal with isolated anomalies and speckle anomalies. First, use mean filtering to process isolated anomalies of electrical performance test data, then use wavelet transform to process speckle anomalies, and finally recover the zero depth level. Simulation results prove the effectiveness and feasibility of the combined filtering algorithm, which provides a new idea for data processing of electrical performance.