Heterologous image target detection method based on intelligent evolution, storage medium and equipment

Abstract

The invention discloses a heterologous image target detection method based on intelligent evolution, a storage medium and equipment, and the method comprises the steps: building a to-be-matched data set step by step in a staged manner based on a deep convolutional neural network; based on a neural network algorithm D2-net, establishing a deep learning network model by utilizing the data set; carrying out training and transfer learning on the established deep learning network model; and carrying out feature point extraction on the heterogeneous images by using the neural network model obtainedthrough transfer learning, and selecting matched targets in the heterogeneous images by using inner point boxes in the feature points so as to complete target detection. According to the intelligent method training framework provided by the invention, continuous evolution of an algorithm model can be completed, and cross-source target matching can be accurately completed.

Description

technical field [0001] The invention belongs to the technical field of target detection, and in particular relates to an intelligent evolution-based heterogeneous image target detection method, storage medium and equipment. Background technique [0002] At present, with the wide application of image matching technology, its research has achieved remarkable results. Among them, feature space and similarity measure criteria are the key elements to realize image matching. The choice of feature space determines the amount of features involved in matching; the similarity measure refers to what is used to determine the similarity between the features to be matched, which is usually in the form of a cost function or a distance function. Similarity measures are closely related to the selection of feature spaces, and they are the key to determine the matching accuracy. Usually, if the matching features used are determined, the similarity measurement criterion is also determined acc...

AI Technical Summary

Problems solved by technology

At present, most heterogeneous image target matching methods are based on traditional methods of image feature extraction and feature point matching. , so it is easy to have a large number of mismatched feature points, so the reliability of this method is relatively poor, and it cannot solve the problem of heterogeneous image target matching with drastic changes in image content.

Images