This is a step by step visualization of Exposure Fusion technique by Mertens et all in popular paper Exposure Fusion which is used for generating HDR images by merging multiple images captured at different exposures.

First for each image a weight map is computed based on objective parameters like Contrast, Saturation & Well exposedness. The weights are computed as follows:

1. Contrast we apply a Laplacian filter to the grayscale version of each image, and take the absolute value of the filter response. This yields a simple indicator C for contrast.
2. Saturation We include a saturation measure S, which is computed as the standard deviation within the R, G and B channel, at each pixel
3. Well-exposedness We want to keep intensities that are not near zero (underexposed) or one (overexposed). We measure this based on closeless to an average intensity of 122 (for 8bit values). This is measured using gaussian curve:
   

The weights are then merged to create a single weight map using

Then the weights are normalized to fall under range [0, 1] using

Contrast map skipped for now. TODO(mebjas) fix this.

Technique by Burt and Adelson [2] is used to blend the input and weight maps into single fused HDR image. The original technique was proposed to blend tow images guided by an alpha mask and works at multiple resolution usinmg a pyramidal image decomposition.

1. This visualizer is built using OpenCV js - https://docs.opencv.org/3.4.0/opencv.js
2. Laplacian of image - https://docs.opencv.org/3.4/da/d85/tutorial_js_gradients.html
3. Saturation Map - Code reference
4. Exposure Map - Code reference

1. Exposure Fusion by Tom Mertens, Jan Kautz, Frank Van Reeth
2. P. Burt and T. Adelson. The Laplacian Pyramid as a Compact Image Code. IEEE Transactions on Communication, COM-31:532–540, 1983