Advertising images play a key role in digital marketing and are an indispensable tool to attract users, convey messages and promote products. However, current advertising image generation faces many challenges, including insufficient image quality and diversity, and high production costs. In this context, the rise of deep learning technology provides a powerful tool to solve these problems. Deep learning is increasingly used in the field of computer vision, bringing new hope for advertising image generation. This article introduces an innovative Transformer-cGANs model that combines the sequence modeling capabilities of Transformer with the image generation technology of Generative Adversarial Networks (GANs). This model shows excellent performance in advertising image generation tasks in digital marketing. Through extensive experimental verification, our model successfully overcomes the insufficient image quality and diversity issues present in traditional methods.