Based on the hyperspectral imaging technology, a non-destructive detection method for pesticide residues on fruit surfaces is proposed. By preprocessing the collected data and extract features, finding the optimal least squares support vector machine parameters through the bacterial population chemotaxis algorithm, a pesticide residue detection model was established, which was compared with the least squares support vector machine model to verify the superiority and accuracy the model. The results showed that the detection model had the highest detection accuracy based on the characteristic wavelength of the continuous projection method combined with the detection model, and its accuracy rate was 97.92%, which had certain application value.