Sum Heatmap Images
Generates an image using pixelwise addition of the selected frames. Processing steps:
-
Background Reference Creation: Performs an iteration over the frames to calculate the Average Z-projection of the raw footage.
-
Background Subtraction: Starts a second iteration over the frames. In this step, subtracts the generated average background from each frame. (This step is implemented by adding the inverted background reference to each frame)
-
Brightness Adjustment: Increases the overall brightness.
-
Frame Sum: Ending the second frame iteration, this operation sums all pixel values into a single 32-bit frame.
-
Final Composition: The sum is inverterd and a simple B&C adjust is applied, stretching the histogram to its boundaries. After that, the selected Lookup table (LUT) is applied.
Interface

Expected Output

Input Video
Any videos supported by ffmpeg. This includes most video formats in existence. Notably, one exclusion: AVI videos saved from ImageJ with compression set to "None" (codec rawvideo) will not work.
This function expects a video with dark subjects over a bright background. If that does not correspond to your data, videos can be inverted using FFmpeg's Invert Colors option or the Invert Video tool.
Output Folder
Where output images will be saved with .tif format.
Lookup Table
The LUTs available are the ImageJ's built-in ones (found at Image → Lookup Tables).
Brightness Tolerance (%)
Percentage of the ROI pixels allowed to sit above the brightness upper bound.
The brightness adjustment stretches the accumulated signal over the available range so the result is not dominated by the darkest and brightest pixels inside the ROI. Bright pixels, however, are often single saturated spots: a sensor hit, a reflection, or a compressed video artifact. When the brightest pixel of the ROI is one of those, it becomes the upper bound and every other pixel gets squeezed towards black, which is what makes the adjustment appear to do nothing.
This parameter sets an upper bound that ignores that kind of outlier. The bound becomes the
(100 - tolerance)-th percentile of the ROI pixels, so 0.01 means "the 99.99th percentile":
only the 0.01% brightest pixels of the ROI are disregarded. Since the percentile is taken over
the ROI itself, the number of tolerated pixels scales with the ROI size, from a handful of
pixels in a small ROI to a few hundreds in a large one.
Pixels above the computed bound are not discarded: they saturate at the top of the range, exactly like a regular brightness and contrast upper limit would.
0 disables the tolerance and uses the plain ROI maximum.
Intervals
The checkboxes either activate or deactivate the relative interval.
Frame numbers should be presented as initialFrame-endFrame, both being inclusive, like ImageJ's built-in Image → Duplicate feature.
An empty interval field results in a skipped interval, while a malformed one (without a - separator) throws an error.