As die size and complexity increase,accurate and efficient extraction of the critical area is essential for yield prediction. Aiming at eliminating the potential integration errors of the traditional shape shifting method,an improved shape shifting method is proposed for Manhattan layouts. By mathematical analyses of the relevance of critical areas to defect sizes,the critical area for all defect sizes is modeled as a piecewise quadratic polynomial function of defect size,which can be obtained by extracting critical area for some certain defect sizes. Because the improved method calculates critical areas for all defect sizes instead of several discrete values with traditional shape shifting method,it eliminates the integration error of the average critical area. Experiments on industrial layouts show that the improved shape shifting method can improve the accuracy of the average critical area calculation by 24.3% or reduce about 59.7% computational expense compared with the traditional method. A die at eliminating the potential integration errors of the traditional shape shifting method, an improved shape shifting method is proposed for Manhattan layouts. By mathematical analyzes of the relevance of critical areas to defect sizes, the critical area for all defect sizes is modeled as a piecewise quadratic polynomial function of defect size, which can be obtained by extracting critical area for some certain defect sizes. all defect sizes instead of several discrete values ​​with traditional shape shifting method, it eliminates the integration error of the average critical area. Experiments on industrial layouts show that the improved shape shifting method can improve the accuracy of the average critical area calculation by 24.3% or reduce about 59.7% computational expense compared with the traditional method.