This paper proposes a reserved quantization indices method for saturated measurements in compressed sensing. The existing approaches tailored for saturation effect do not provide a way to identify saturated measurements, which is mandatory in practical implementations.We introduce a method using reserved quantization indices to mark saturated measurements, which is applicable to current quantizer models. Two extended approaches based on the proposed method have been investigated compared to the existing approaches. The investigation shows that saturated measurements can be identified by reserved quantization indices without adding extra hardware resources while maintaining a comparable reconstruction quality to the existing approaches.
Proceedings of the 12th Ieee International Symposium on Signal Processing and Information Technology, 2012, p. 198-202
IEEE International Symposium on Signal Processing and Information Technology, 2012