Fig. 4: Working principle of snapshot compressive imaging (SCI) enabled by PPMS. | Nature Communications

Fig. 4: Working principle of snapshot compressive imaging (SCI) enabled by PPMS.

From: In-sensor compressing via programmable optoelectronic sensors based on van der Waals heterostructures for intelligent machine vision

Fig. 4

a Demonstration of the optoelectronic AND logic gate for sensing and compressing. b An algorithm flow chart for encoding and compressing dynamic images by using the sensor array. Lij and Eij denote the optical signal and electrical signal applied to the sensor, respectively. Xk and Mk correspond to the pixel array of kth frame (tk) and kth mask, respectively. Cij corresponds to the compressed value at pixel(i,j), mapped to the drain-source current of the channel after 2D measurement with the final output matrix denoted as Y (Ids, Y). c Current evolution in the sensor under different stimulation modes: Electrical-only (blue bar), optical-only (green bar), and electro-optical co-stimulus (red bar). d The sensor can be finely programmed with electro-optical co-stimulus, showing linear decrease with 19 levels. Inset: linear fit between the channel current in each conductance state and the pulse number. The enlarged electro-optical response of the sensor under different lasers: (e) 532 nm (0.018 nW/1 s), (f) 450 nm (0.0085 nW/1 s), and (g) 635 nm (0.018 nW/1 s). The blue dashed lines indicate the transient current variation triggered by electrical pulses.

Back to article page