For small sample targets,a method for the implementation of small sample images two-dimensional to stereo view based on experience valueξfactor was proposed.Firstly,the original feature points of small samples were extracted,and the transformation arrays of projection,rotation,scaling and translation were constructed based on the quadratic element methods.Secondly,construction factor rotation transformation quantity and translation transformation quantity based on small samples were obtained.Thirdly,based on the transformation,the plane points were inversely mapped into three dimensional space points.Fourthly,the experience valueξfactor was set and the surface patch was created by infinite approximation.Finally,the optimal goal was constructed by iterating.The experimental results showed that the proposed method had low computational cost,high efficiency and good visibility of reconstructed view for small samples.
关键词
小样本/射影/自标定/贴片/经验值/稀疏稠密化
Key words
small samples/projective/self-calibration/paster patch/experience value/sparse densification
维普
万方数据