The personalization-based movie recommendation necessitates precise modeling of the characteristics of both the user and the movies. Conventional methods which primarily depend on visual perception using posters or frames are not efficient in gaining the semantic information embedded in the images. To overcome this shortcoming, we develop a hybrid quantum classical deep learning system comprising of Variational Quantum Circuits (VQCs) and a visual module based on yolo. The YOLO component identifies important features in movie posters and frames, i.e. actors, objects, and scenes and user interaction data is included to add layers of understanding. We have also added these multimodal characteristics to a quantum-enhanced model that takes advantage of VQCs to effectively model complex and high-dimensional user-style relationships. Experiments on benchmark datasets show that the proposed system can enhance recommendation accuracy and personalization in comparison to those of traditional classifiers. What is more, our approach demonstrates higher performance when compared to baseline models, such as Support Vector Machines (SVM) and Decision Trees, with the accuracy of 96%. This paper shows the possibility of using quantum computing and deep learning and computer vision to improve the next generation of intelligent movie recommendation systems.