1. Image Histogram Analysis
- Walter Cheng
- May 30, 2019
- 2 min read
Updated: Jun 27, 2019
In order to generate a science-fiction style image, we would like to take a look at the idea of this kind of image. One of the way to analysis the information of one image, is to investigate the image histogram.
Image histogram is a kind of graph that represents the pixel information of an image, from 0 to 255. The more bright pixels the image has, the histogram will have a larger value in the right hand side.
In order to generate image histogram, we will first convert the image into gray scale, and then calculate how many pixels are there for each value.
We would use this image as an example for generating an image histogram:
As we can analysis from most science-fiction style image, it seems like the light is brighter than a normal image, and the shadow is darker than a normal one. Therefore, we should expect a U-shape for the image histogram for science-fiction style image.
In here, we would use this as an example:
As we can observe, the image histogram is an obvious U-shape since most pixels are focused on the very dark and very light colors. However, the more images that we tried to test, it doesn't seem like it is always the cases.
It might seems like there are some U-shape within those science-fiction style image, but it is not obvious. However, there are still one thing we could observe in this similar image.
The fourth sample and the third sample picture seems to have the similar percent amount of light in the image, and as we can see, the image histogram for two of these images look similar.
Therefore, although we cannot be sure about the shape or the characteristic in the image histogram of the science-fiction style image, we can still spot the similarities within the similar style.
With those information, we can make good use of these kinds of sample images and use those to generate the filter effect that we want. In order to achieve that, we can use histogram matching to force an image to have the same image histogram as the sample image without losing any important information. These will be discussed in the next chapter.
Image credit:
Comments