With its advantages of wide breadth, strong comprehensive coverage and short revisit period, GF-1 satellite has become one of the important data sources used in land and resources survey, remote sensing monitoring of agriculture and forestry, and major national engineering construction. In this study, the authors took the China-Pakistan Economic Corridor as an example, selected the GF-1 WFV data for comparative experiments, and analyzed two key factors that affect the application of remote sensing images, i.e., how to improve the geometric positioning accuracy of remote sensing images and the true color image synthesis method. The experimental results indicate that the block adjustment model based on rational function model technology and RGB-NIR color synthesis model have good performance in image accuracy and visual effect of imagery respectively. The X-direction residual and Y-direction residual of ortho-image were improved to 0.79 pixels and 0.83 pixels. The information entropy, average gradient, mean and standard deviation of the resulting images were improved in varying degrees. This method could not only ensure the true and natural color of the image but also keep the information details, with the image surface effect better optimized. This test method is a better strategy for large scale data application in actual production.