Hypersonic internal compression inlets are typically designed using streamline tracing technology based on the basic flowfield, and the quality of the basic flowfield design directly determines the performance metrics of the inlet. In this study, the quasi-uniform B-spline method is applied to achieve the parametric design of the internal compression basic flowfield. A fast prediction model for the internal compression basic flowfield is established based on the residual neural network architecture of deep learning, realizing the goal of "parametric design-flowfield prediction." The predicted flowfield cloud images are quantitatively evaluated using image quality assessment methods, and key flowfield characteristic parameter distributions are extracted to enable rapid acquisition of flowfield cloud images and characteristic parameter distributions based on design parameters. The results show that the developed basic flowfield fast prediction model achieves high prediction accuracy for flowfields corresponding to given geometric parameters. The average peak signal-to-noise ratio (PSNR) of the predicted total flowfield is 42.51 dB, and the average structural similarity index (SSIM) is 0.9973. Additionally, the model can effectively extract key flowfield characteristic parameter distributions from the predicted results, providing robust support for the rapid design and optimization of internal compression basic flowfield.