Color Image Retrieval Method Using Low Dimensional Salient Visual Feature Descriptors for IoT Applications

Table 5

Comparative results in terms of average precision for top 20 retrieved images for Corel-1K dataset with some state of art CBIR schemes.

Image category	Shrivastava and Tyagi [24]	Dubey et al. [25]	Mistry et al. [26]	Irtaza et al. [27]	Ahmed et al. [28]	Garg and Dhiman [30]	Vimina and Divya [29]	Varish and Singh [31]	Kumar et al. [32]	Joseph et al. [33]	Proposed scheme
People	74.80	75.00	62.00	83.00	90.00	81.20	72.50	60.00	80.00	81.00	91.95
Beaches	58.20	55.00	71.00	72.00	60.00	96.56	52.75	55.00	65.00	78.00	77.05
Monuments	62.10	67.00	53.00	86.00	90.00	78.20	52.70	50.00	65.00	80.00	83.75
Buses	80.20	95.00	77.00	100.00	75.00	83.30	93.95	95.00	100.00	81.00	93.55
Dinosaurs	100.00	97.00	89.00	97.00	100.00	82.20	99.45	100.00	100.00	100.00	95.25
Elephants	75.10	63.00	55.00	82.00	70.00	77.50	52.40	45.00	70.00	81.00	84.15
Flowers	92.30	93.00	89.00	86.00	90.00	86.30	86.65	70.00	95.00	89.00	94.90
Horses	89.60	89.00	55.00	82.00	100.00	86.10	83.75	95.00	100.00	95.00	90.35
Mountains	56.10	45.00	52.00	69.00	70.00	86.30	43.15	80.00	70.00	68.00	72.70
Foods	80.30	70.00	55.00	90.00	90.00	90.20	71.05	65.00	90.00	60.00	85.20
Mean	76.90	74.90	65.80	84.70	83.50	84.79	71.04	71.50	83.50	81.30	86.89

The bold value indicates the mean representation.