The development of digital image processing technology enables automatic object identification with high accuracy. This study aims to classify images of musical instruments, namely maracas, guitars, and drums, using a combination of K-Means-based color segmentation and Gray Level Co-Occurrence Matrix (GLCM) feature extraction. The process begins with converting RGB images into the Lab color space, followed by object segmentation using the K-Means clustering algorithm to separate the main object from the background. Subsequently, shape features (metric, eccentricity) and texture features (contrast, correlation, energy, homogeneity) are extracted using GLCM. The extracted features are then compared with a feature database using a distance-based approach to determine the object class. Experimental results show that the system can successfully recognize maracas, guitar, and drum images with a satisfactory accuracy level. This research demonstrates that the combination of K-Means and GLCM methods can serve as an effective approach for musical instrument image classification and has the potential to be further developed for object recognition in other fields