This paper collects the lighting color images in the musical scene with the help of the panoramic camera, constructs a color evaluation function with the color harmony factor, color emotion factor, and color contrast factor as the main components, uses the genetic algorithm in the optimization algorithm to optimize the above color evaluation function, and then completes the design of the music lighting color mixing template. Comprehensive research data and calculation formulas are used to verify the optimization model of musical theater performance effects based on constructivism and genetic algorithms. Among the six color emotion factors, the red emotion factor is the most prominent with a score of 579, and the corresponding emotional characteristics are passion, love, vitality, and positivity. Meanwhile, the overall stage lighting intensity of the musical Cats was measured to be 30.9361. In addition, the adaptation value of the color evaluation function based on the genetic algorithm was 0.897. The Musi CNN model’s classification of music labels yielded a correct rate of 97.58% for both the music label training and test sets.