In this paper, we present a fast and robust color feature extraction for effective content-based image retrieval (CBIR). In color feature extraction, since cylindrical HSV color space is not perceptually uniform, we improve the color quantization and similarity measure method based on a cylinder-cone transform. In image segmentation, speed is more important than accuracy in CBIR. We propose a new rectangular approximate image segmentation to solve the problem. We also develop a significance function to reflect the importance of different position in image, and improve the segmentation and retrieval performance. Finally, we present a similarity measure between images with multi-objects. Experimental results show that our proposed method outperforms the existing method in terms of efficiency and robustness.