In the current booming digital media, traditional image segmentation methods in digital media applications generally have problems such as low segmentation accuracy, poor interactivity, low algorithm efficiency, and excessive memory usage. Based on this, this paper introduces a digital media deep interactive genetic algorithm based on VR intelligent image segmentation. First, this paper uses VR technology to build a three-dimensional image interactive environment, applies the genetic algorithm to the image segmentation process, and iterates and optimizes the image segmentation scheme through genetic operations. In the implementation of the genetic algorithm, binary coding is used to represent the image segmentation scheme, and the fitness function is used to evaluate the segmentation effect. During the execution of the algorithm, VR interaction is combined with genetic algorithm iteration. Users can view and adjust the preliminary segmentation results in real time in the VR environment, and these adjustments will serve as feedback information to guide the next iteration of the genetic algorithm. The interactive genetic algorithm proposed in this paper is superior to the single algorithm in segmentation accuracy, with an average segmentation accuracy of up to 94.62%, while the single algorithm is about 81.75%. During the execution of the algorithm, VR interaction is combined with genetic algorithm iteration. Users can view and adjust the preliminary segmentation results in real time in the VR environment, and these adjustments will serve as feedback information to guide the next iteration of the genetic algorithm. The interactive genetic algorithm proposed in this paper is superior to the single algorithm in segmentation accuracy, with an average segmentation accuracy of up to 94.62%, while the single algorithm is about 81.75%.