Research on Neural Network Model Based on an Intersection-Free Segmentation Algorithm and HSV Feature Classification

This paper innovatively creates a color recognition and classification method, which leverages an intersection-free segmentation algorithm (IFSA). The IFSA enables precise segmentation and deduplication of multiple objects, effectively reducing detection errors typically caused by overlapping. After applying IFSA and image preprocessing steps, a neural network model then classifies color categories with an accuracy of 98.75%. Experimental results highlight IFSA's superior accuracy and stability, particularly in overlapping scenarios, offering a robust solution for multi-target color recognition.