Introduction: Accessible movies, which primarily employ descriptive video as their presentation method, serve as a form of cultural exchange without barriers. They have built a bridge for individuals with visual impairments (ie, those who are blind or have low vision) to engage with film and television culture and to communicate with society. This study developed a theoretical model for audio description creation by coding 10 accessible movies, addressing issues of inconsistency and lack of standardization, to enhance the viewing experience for individuals with visual impairment. Methods: The audio description of 10 representative accessible movies was selected and transcribed into text. Using grounded theory analysis and the qualitative coding tool Nvivo12 Plus, the texts underwent three-level coding, resulting in a theoretical model for the audio description creation process. Results: The analysis identified two major themes: text deconstruction and meaning reconstruction, leading to the concept of “from de-alienation to re-alienation.” The model explained the process of conveying films to individuals with visual impairments by first deconstructing the work to facilitate basic plot comprehension, and then reconstructing the work's meaning to enhance character images, supplement and interpret plots, and summarize significance. Discussion: The model addresses how creators should restore basic storylines and reconstruct deeper meanings, improving the quality of audio descriptions. Implications for Practitioners: This model provides new insights into the audio description creation process in China, promoting greater standardization and helping creators produce meaningful and reflective content for viewers with visual impairments.