Microwave photonic radars can transmit large bandwidth and high carrier frequency signals, which makes two-dimensional high-resolution Inverse Synthetic Aperture Radar (ISAR) imaging possible. It is important to study the corresponding real-time imaging algorithms. However, the high range resolution and high carrier frequency of the signal make the space curvature of the distance bend non-negligible. This is the reason for the poor imaging performance of the traditional Doppler real-time imaging algorithm. In addition, the computationally intensive beam domain imaging algorithm is not suitable for microwave photonic radar signals of large data volume. Therefore, a high-efficiency microwave photonic ISAR high-resolution real-time imaging algorithm is proposed in this paper. Firstly, this algorithm uses the Generalized Keystone Transform (GKT) to extract the phase of the special display point. Next, it inverts the target lateral velocity from phase modulation frequency. Finally, the result of velocity estimation and Frequency Scaling (FS) are used to correct the distance space-bending and conduct matched filtering imaging in azimuth. The simulation results and the measured data have been shown to verify the effectiveness of the proposed algorithm.