xmipp3.protocols.protocol_preprocess.protocol_movie_resize module
- class xmipp3.protocols.protocol_preprocess.protocol_movie_resize.XmippProtMovieResize(**args)[source]
Bases:
ProtProcessMoviesResize a set of movies. Only downsampling is allowed.
AI Generated
## Overview
The Movie Resize protocol downsamples a set of cryo-EM movies.
Movies can be very large, and many early processing or quality-control steps do not require the original full pixel size. Downsampling movies reduces file size, memory use, and computational cost. It can also prepare movies for workflows that expect a specific sampling rate or image dimension.
This protocol resizes each movie in Fourier space and produces a new set of movies with updated sampling rate metadata.
Only downsampling is allowed. The protocol does not support increasing movie dimensions or decreasing the sampling rate.
## Inputs and General Workflow
The input is a set of movies.
For each movie, the protocol computes the new movie dimensions and output sampling rate according to the selected resize option. It then runs xmipp_image_resize on the movie stack, preserving the number of frames while changing the X and Y image size.
The output movies are written as MRC movie stacks with names such as:
movie_000001_resize.mrcs
The protocol supports streaming. As new movies arrive in the input set, they are resized and appended to the output movie set.
## Input Movies
The Input movies parameter defines the movie set to be resized.
All movies are expected to share the dimensions and sampling rate stored in the input movie set. The protocol uses those values to compute the output size and sampling rate.
The input movies are not modified. The protocol writes new resized movie files.
## Resize Option
The Resize option parameter defines how the downsampling is specified.
There are three options:
Sampling Rate: the user provides the desired output sampling rate.
Dimensions: the user provides the desired output image size in pixels.
Factor: the user provides a downsampling factor.
In all cases, the protocol computes a new X and Y dimension and keeps the same number of movie frames.
## Resize by Sampling Rate
When Sampling Rate is selected, the user provides the desired output sampling rate in angstroms per pixel.
The requested sampling rate must be larger than the original sampling rate, because only downsampling is allowed.
The protocol computes the downsampling factor as:
- [
ext{factor} =
- rac{ ext{new sampling rate}}{ ext{original sampling rate}}
]
and then computes the new image dimension from that factor.
This option is useful when the user wants the output movies to match a specific physical pixel size.
## Resize by Dimensions
When Dimensions is selected, the user provides the desired output image size in pixels.
The new dimension must be smaller than or equal to the original movie dimension, because only downsampling is allowed.
The protocol computes the new sampling rate from the ratio between the original and new dimensions.
This option is useful when a downstream workflow requires a specific movie width and height.
## Resize by Factor
When Factor is selected, the user provides the downsampling factor.
The factor must be greater than or equal to 1. A factor of 2 halves the image dimensions and doubles the sampling rate. A factor of 4 reduces the dimensions by four and multiplies the sampling rate by four.
This option is useful when the user wants a simple integer or numerical downsampling ratio.
## Fourier Resize
The protocol uses Fourier resizing.
For each movie, it calls xmipp_image_resize with the Fourier option and the new X, Y, and frame dimensions. The number of frames is kept unchanged.
Fourier resizing is appropriate for downsampling because it changes the spatial sampling while preserving the frequency-domain interpretation of the movie data.
## Output Movies
The main output is outputMovies.
This output is a new Scipion movie set containing the resized movies. Each output movie:
keeps the same object identifier as the corresponding input movie;
points to the resized .mrcs file;
copies the acquisition information;
stores the updated sampling rate;
keeps the same frame range as the input movie set.
The output set can be used in downstream movie-processing workflows.
## Streaming Behavior
The protocol supports streaming input.
As new movies appear in the input set, the protocol inserts processing steps for them. Each movie is resized independently. When a movie has been processed, it is appended to the output movie set.
The output set remains open while the input stream is open. When the input stream closes and all movies have been processed, the output stream is closed.
This makes the protocol suitable for online processing workflows where movies arrive progressively during data acquisition.
## Summary Information
The protocol summary reports the number of input movies, their original size, the number of output movies, and the resized output size.
This allows the user to quickly verify that the expected downsampling was performed.
## Validation Rules
The protocol validates that only downsampling is requested.
For Factor, the factor must be greater than or equal to 1.
For Dimensions, the requested new dimension must not be larger than the input movie dimension.
For Sampling Rate, the requested sampling rate must be larger than the original sampling rate.
If these conditions are not met, the protocol reports an error.
## Interpreting the Result
The resized movies should be interpreted as downsampled versions of the input movies.
The protocol does not align, dose-weight, correct motion, estimate CTF, or change the number of frames. It only changes the spatial sampling of each movie.
Downsampling reduces high-resolution information. This is often acceptable for early quality-control or low-resolution processing, but it may not be appropriate if the full-resolution movie information is needed later.
## Practical Recommendations
Use this protocol to reduce movie size and accelerate early processing.
Choose resizing by sampling rate when the output should match a target pixel size.
Choose resizing by dimensions when a downstream protocol expects a specific image size.
Choose resizing by factor for simple downsampling ratios.
Do not use downsampled movies for final high-resolution workflows unless the chosen sampling rate still supports the target resolution.
Keep the original movies available if later full-resolution processing may be needed.
In streaming workflows, use the output movie set as the downsampled stream for downstream protocols.
## Final Perspective
Movie Resize is a movie-preprocessing protocol for spatial downsampling.
For biological users, its value is that it creates smaller and faster-to-process movie stacks while preserving movie metadata such as acquisition information, frame range, and updated sampling rate.
The protocol is useful for reducing computational cost and preparing movies for early processing, but it should be used with awareness that downsampling limits the highest resolution that can be recovered from the resized data.