Grayscale and color ghost-imaging of moving objects by memory-enabled, memoryless and compressive sensing algorithms

Grayscale and color computational ghost imaging of a moving object is investigated. Three different ghost imaging algorithms, memory-enabled, memoryless and compressive sensing, are employed and compared. A speckle resizing method is also applied to enhance the features of the obtained ghost images. The results indicate that the compressive sensing algorithm provides better SNR and visibility values, while the memory-enabled algorithm offers higher imaging speeds. It is shown that for shot numbers ranging from 150–400, SNR values as high as 1.5, visibilities around 0.95 and maximum target speeds about 0.65 pixel/s are affordable by proper selection of the applied ghost imaging algorithm. Also, a method for color ghost imaging of moving objects through RGB illumination is introduced. Furthermore, results for experimental computational ghost imaging of a moving object along with the maximum practicable speed are included. (paper)