Animation Scene Object Recognition and Modeling Based on Computer Vision Technology

Based on computer vision technology, this article proposes a computer-aided design (CAD) modeling method for animation scene objects based on recognition. It conducts a series of experiments to verify the effectiveness and practicality of this method. This method aims to automatically and accurately recognize objects from animation scenes and generate corresponding CAD models using object recognition algorithms and CAD modeling techniques. To achieve this goal, the article first uses the Convolutional Neural Network (CNN) algorithm for object recognition and obtains the boundary information of the object. Then, CAD modeling uses boundary information based on recognition to generate CAD models similar to real objects. The performance of the system was assessed through a series of experiments. The results show that this method has a high accuracy in object recognition, with an essential accuracy of over 93%, and can accurately extract boundary information of objects. The gap between the generated CAD model and the real object is small in CAD modeling, and the modeling accuracy is high. Moreover, the system's running time is within an acceptable range, proving the system's efficiency. This method can provide animators with an automated CAD modeling tool, reducing the time and effort costs of manual modeling and improving modeling accuracy. © 2024 U-turn Press LLC.