How Can We Achieve Fast Bitmap Blurring in Android?

Fast Bitmap Blur for Android

In Android, applying a blur filter to images can be a time-consuming process, especially for large images. This is because traditional methods involve iterating through each pixel of the image, which can be computationally expensive.

Introducing a Faster Algorithm

The article presents a faster algorithm than the traditional pixel-by-pixel blurring approach. This algorithm, ported from Quasimondo, combines elements of both box blur and Gaussian blur.

The algorithm works by creating a "moving stack" of colors as it scans through the image. This stack is updatedincrementally, with a new color added to the right side and the leftmost color removed. The remaining colors on top of the stack are adjusted depending on their position on the stack.

Implementation Details

The provided Java code snippet implements the algorithm in Android. It takes three parameters:

• sentBitmap: The input bitmap to be blurred
• scale: The scale factor to reduce the image size before applying the blur
• radius: The blur radius to be applied

The algorithm is divided into two nested loops, one for rows and one for columns. Within each loop, the algorithm calculates the summations of colors and updates them using the moving stack approach.

Performance Improvement

This algorithm significantly improves blurring performance compared to pixel-by-pixel methods. For a 640x480 image, the original algorithm took 30 seconds to blur, while this algorithm reduces that time to around 5 seconds.

Note

The code snippet includes a fix for an ArrayIndexOutOfBoundsException issue that may occur in certain cases. For Android 11 and later, ensure you use StrictMath.abs instead of Math.abs to prevent the exception.

The above is the detailed content of How Can We Achieve Fast Bitmap Blurring in Android?. For more information, please follow other related articles on the PHP Chinese website!