A new approach for straightness fitting of micro-drill’s main lips is proposed. While its chips are measured, micro-drill’s projective images are collected with high precision automatic test system of PCB Micro-drill, and sub-pixel edge features are extracted. The sum of square distances' from the coordinates of sample points to fitted line is taken as objective function. Then a Sanger neural network with lateral connection is designed, where a self-adaptive minor component extracting method is adopted. When the system comes to equilibrium, the eigenvector of minimum eigen value is the fitted line coefficient, from which chips of micro-drill main lips are obtained. The proposed approach is the novel application of Sanger neural network in straightness fitting.