You are here

A note about BMV image frame data

Some map challenge participants may be encountering difficulty processing the BMV individual frames data (not the original boxed out particle images). One has to work out the image processing scheme which is a part of the challenge. To help challengers understand the data, the BMV raw data providers have provided the following note:



Note on the BMV image frame data -- provided by Benjamin Bammes (Wang, Z, Hryc, CF, Bammes, B, Afonine, PV, Jakana, J, Chen, DH, Liu, X, Baker, ML, Kao, C, Ludtke, SJ, Schmid, MF, Adams, PD, & Chiu, W (2014) An atomic model of brome mosaic virus using direct electron detection and real-space optimization. Nature communications 5:4808) :

"We collected dark and gain reference images for each day of data collection except 2013-01-12, for which we used the dark reference image from the following day (2013-01-13). Note that the dark and gain reference images from each day were rotated and/or flipped if necessary to match the orientation of the raw frames from that day.

In order to improve statistics of the gain reference image, we averaged all the dark-subtracted gain reference images to create one average gain reference image to apply to the data from all days of data collection. The average gain reference image was converted to 32-bit float and then divided by its mean intensity, so that the mean intensity of the average gain reference image was normalized to one. We then applied a threshold to the average gain reference image such that all pixel values less than 0.01 were set to 0.01. Finally, we inverted the average gain reference image by calculating the reciprocal value of each pixel.

All raw frames were processed by first applying dark correction, and then applying gain correction. Dark subtraction was performed simply by integer subtraction of the day's dark reference image from each raw frame. Gain correction was then performed by multiplying each frame by the average gain reference image to produce the final flat-field corrected, 32-bit float stacks of frames.

Note that we did not yet apply any sigma filter to remove X-ray pixels and/or detector noise from each frame."

EMDataBank Validation Challenges are supported by NIH National Institute of General Medical Sciences

Please send your challenge questions, comments and feedback to challenges@emdatabank.org

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer