A New Integral Image Structure for Memory Size Reduction

Abstract

SUMMARYAn integral image is the sum of input image pixel values. It is mainly used to speed up the process of a box filter operation, such as Haar-like features. However, large memory for integral image data can be an obstacle in an embedded environment with limited hardware. Therefore, an efficient method to store the integral image is necessary. In this paper, we propose a memory size reduction method for integral image. The method uses four types image information: an integral image, a row integral image, a column integral image, and an input image. Using this method, integral image memory can be reduced by 42.6% on a 640 × 480 8-bit gray-scale input image. The same idea can be applied for bigger size images. key words: integral image, image processing, image recognition, embedded system

A New Method of Storing Integral Image for Memory Efficiency Using Modified Block Structure

“…In a previous research [5], block structure integral image for memory reduction was introduced, but it requires large additional calculations on the block size increases. We transform the equation of the block structure integral image to reduce the number of memory accesses.…”

“…However, large memory capacity for integral image data can be an obstacle in an embedded environment with limited hardware. In a previous research, [5] reduced the size of integral image memory using 2 × 2 block structure with additional calculations. It can be easily extended to n × n block structure for further reduction, but it requires more additional calculations.…”

“…A concept of block structure integral image was introduced in [5] for memory size reduction. In [5], a 2 × 2 size block was used, but the block size can easily be extended.…”

“…In [5], a 2 × 2 size block was used, but the block size can easily be extended. The extended version of the block structure integral image is shown in Fig.…”

“…In a previous research [5], 2 × 2 block structure integral image was introduced to reduced the size of the required memory. The block structure of [5] can be easily extended to a larger size, but it requires more additional calculations as the block size increases. In this paper, we propose a new block structure integral image to reduce the amount of additional calculations, while keeping the same amount of required memory.…”