Holographic image processing with phase error compensation
A method and system of holographic image processing includes phase error compensation.
1 . A method for generating holographic images, the method comprising:
projecting, during calibration, first diffraction pattern images displayed at a spatial light modulator (SLM), the first diffraction pattern images based on first diffraction pattern data, the first diffraction pattern images associated with multiple focal lengths;
generating, during the calibration, model images associated with the multiple focal lengths based on the first diffraction pattern data;
generating, during the calibration, a phase error map based on the model images and captured images of the projected first diffraction pattern images, the captured images associated with the multiple focal lengths; and
adjusting, during run-time, phase values based on the phase error map, the phase values used to generate second diffraction pattern data for one or more second diffraction pattern images.
2 . The method of claim 1 , wherein the model images are non-ideal model images that include a phase error.
3 . The method of claim 1 , wherein the generating of the model images includes inputting a phase profile of the first diffraction pattern data into an optical field propagation model.
4 . The method of claim 1 , wherein the generating of the model images includes inputting the first diffraction pattern data into an optical field propagation model that is based on a convolutional Fresnel diffraction algorithm.
5 . The method of claim 1 , wherein the generating of the model images includes operating a propagation model to generate a model sensor optical field that models an image captured at a sensor array.
6 . The method of claim 5 , including converting the model sensor optical field into an intensity image to form a first one of the model images.
7 . The method of claim 1 , wherein the generating of the phase error map is based on a gradient descent operation, the gradient descent operation based on pixel values of the model images and pixel values of the captured images, the gradient descent operation to generate the phase error map iteratively.
8 . The method of claim 1 , including pre-processing the captured images before generating the phase error map based on the captured images, the pre-processing including vibration compensation and denoising.
9 . The method of claim 1 , wherein the generating of the phase error map is based on a gradient descent operation.
10 . A system comprising:
memory to store holographic image data and phase errors;
instructions; and
at least one processor circuit to be programmed based on the instructions to:
obtain holographic captured images associated with multiple different focal lengths from a spatial light modulator (SLM), the SLM to display diffraction pattern images;
apply vibration compensation to the captured images to generate pre-processed holographic image data, the vibration compensation based on alignment of image content of multiple captured images associated with a same focal length to image content of an anchor image; and
generate a phase error map based on the pre-processed holographic image data and model images, the model images based on data to form the diffraction pattern images.
11 . The system of claim 10 , wherein one or more of the at least one processor circuit is to apply the vibration compensation by adjusting the multiple captured images by an average vibration shift, the average vibration shift based on shifts of the image content in the anchor image relative to the image content of respective ones of the multiple captured images.
12 . The system of claim 11 , wherein one or more of the at least one processor circuit is to apply the vibration compensation by:
determining a tilt mis-alignment shift of a sensor array of a camera relative to the SLM, the camera to provide the captured images; and
adding the tilt mis-alignment shift to the average vibration shift to adjust the image content of the multiple captured images.
13 . The system of claim 10 , wherein one or more of the at least one processor circuit is to average the image content of the multiple captured images at the same focal length to form an average captured image to be used to generate the phase error map.
14 . The system of claim 13 , wherein one or more of the at least one processor circuit is to subtract an average noise estimate from image data of pixel locations of the average captured image.
15 . The system of claim 10 , wherein one or more of the at least one processor circuit is to generate the phase error map based on a gradient descent operation, the gradient descent operation based on the pre-processed holographic image data φ and the model images.
16 . The system of claim 10 , wherein the diffraction pattern images are first diffraction pattern images, and one or more of the at least one processor circuit is to adjust phase values based on the phase error map, the phase values to be used to generate diffraction pattern image data of one or more second diffraction pattern images to be displayed at the SLM.
17 . The system of claim 10 , wherein one or more of the at least one processor circuit is to subtract estimated noise from the captured images.
18 . At least one non-transitory machine readable medium comprising a plurality of instructions to cause at least one processor circuit of a computing device to at least:
cause projection of diffraction pattern images displayed at a spatial light modulator (SLM), the diffraction pattern images based on diffraction pattern data, the diffraction pattern images associated with at multiple focal lengths;
generate model images associated with the multiple focal lengths based on the diffraction pattern data, a model image of the model images generated based on at least one phase error guess input into a propagation model that generates a model sensor optical field, the model sensor optical field convertible into the model image, the phase error guess obtained iteratively from a gradient descent operation;
obtain holographic captured images of the projected diffraction pattern images, the captured images associated with the multiple focal lengths; and
generate a phase error map based on the gradient descent operation, the gradient descent operation based on the model images and the captured images.
19 . The medium of claim 18 , wherein the instructions are to cause one or more of the at least one processor circuit to include a latest phase error guess in the phase error map.
20 . The medium of claim 18 , wherein the gradient descent-type of operation determines phase errors for the phase error map based on an average total minimum mean square error (MSE) between the captured images and the model images at the multiple focal lengths.
21 . The medium of claim 18 , wherein the instructions are to cause one or more of the at least one processor circuit to perform pre-processing on the captured images, the pre-processing including vibration compensation and denoising, the pre-processing performed before the captured images are processed by the gradient descent operation.
22 . The medium of claim 18 , wherein the instructions are to cause one or more of the at least one processor circuit to initialize the gradient descent operation with an initial phase error guess, the initial phase error guess based on a phase profile of an image with a focal length longer than the multiple focal lengths.
23 . The medium of claim 18 , wherein the gradient descent operation modifies at least one of the captured images or the model images based on at least one energy scaling factor.
24 . The medium of claim 18 , wherein the instructions are to cause one or more of the at least one processor circuit to generate the model images based on SLM parameters that generate a phase profile with no phase error.