Space target geometric constraint auxiliary robust detection method based on multiple view fields

The invention provides a multi-view-field-based space target geometric constraint auxiliary robust detection method, and relates to the field of space target detection, and the method comprises the steps: building a multi-view-field-based target detection model, and extracting a multi-view-field image sequence according to the target detection model; the method comprises the following steps: extracting two-dimensional point features and line features based on a multi-view-field image sequence, calculating a corresponding three-dimensional line set through trifocal tensor screening and matching, estimating a plane set according to the three-dimensional line set, calculating a corresponding homography matrix, mapping points among images through the homography matrix, and establishing a matching relationship; based on the matching relation, managing the candidate points to obtain mature points, calculating GDOP weighting factors, using the mature points and the GDOP weighting factors for BA optimization, and cal