PRUDA: A Novel Measurement Attribute Set towards Robust Steganography in Social Networks

Table 1

Nonuniform measurement attributes of existing methods.

Attack type	Method	Payload	Robustness	Ease of Use	AntiDetection	Applicability
Scaling	Method in [27]	0.1–0.5bpp of scaled image		Any image	0.4936	Known scaling factor and scaling type is the nearest-neighbor interpolation
	Method in [29]	96bit		Any image	0.2082	Scaling factor greater than 0.5
	Method in [30]	128bit		Any image	0.3395	Scaling factor greater than 0.5
Compression	Method in [15]	0.1–0.15bpnC of original image		Repeatedly compressed image	0.405	No restriction
	JCRISBE [12]	0.05–0.3bpnAC of original image (BCH)		Repeatedly compressed image that meets certain requirements	0.43	Known accurate channel compression factor
	Method in [13]	0.1–0.5bpnAC of compressed image		Any image	0.42	Known accurate channel compression factor
	GMAS [16]	0.05–0.15bpnAC of precompressed image (RS)		The precompressed image	0.025	Known approximate channel compression range
	Method in [14]	0.05–0.35bpnAC of original image (BCH)		Any image	0.42	No restriction
	DCRAS [8]	0.01–0.1bpnAC of precompressed image (RS)		The precompressed image	0.055	Known approximate channel compression range
	Method in [31]	0.01–0.1bpnAC of precompressed image (RS)		The precompressed image	0.0227	Known approximate channel compression range
Multiple attacks	MREAS- [17]	0.01–0.1bpnAC of precompressed image (RS)		The image that conforms to certain laws	Close to zero	Known approximate channel compression range