Integral Images: Efficient Algorithms for Their Computation and Storage in Resource-Constrained Embedded Vision Systems.
Bottom Line:
The integral image, an intermediate image representation, has found extensive use in multi-scale local feature detection algorithms, such as Speeded-Up Robust Features (SURF), allowing fast computation of rectangular features at constant speed, independent of filter size.An efficient design strategy is also proposed for a parallel integral image computation unit to reduce the size of the required internal memory (nearly 35% for common HD video).Addressing the storage problem of integral image in embedded vision systems, the paper presents two algorithms which allow substantial decrease (at least 44.44%) in the memory requirements.
View Article:
PubMed Central - PubMed
Affiliation: School of Computer Science and Electronic Engineering, University of Essex, Colchester CO4 3SQ, UK. sehsan@essex.ac.uk.
ABSTRACT
The integral image, an intermediate image representation, has found extensive use in multi-scale local feature detection algorithms, such as Speeded-Up Robust Features (SURF), allowing fast computation of rectangular features at constant speed, independent of filter size. For resource-constrained real-time embedded vision systems, computation and storage of integral image presents several design challenges due to strict timing and hardware limitations. Although calculation of the integral image only consists of simple addition operations, the total number of operations is large owing to the generally large size of image data. Recursive equations allow substantial decrease in the number of operations but require calculation in a serial fashion. This paper presents two new hardware algorithms that are based on the decomposition of these recursive equations, allowing calculation of up to four integral image values in a row-parallel way without significantly increasing the number of operations. An efficient design strategy is also proposed for a parallel integral image computation unit to reduce the size of the required internal memory (nearly 35% for common HD video). Addressing the storage problem of integral image in embedded vision systems, the paper presents two algorithms which allow substantial decrease (at least 44.44%) in the memory requirements. Finally, the paper provides a case study that highlights the utility of the proposed architectures in embedded vision systems. No MeSH data available. |
Related In:
Results -
Collection
License getmorefigures.php?uid=PMC4541907&req=5
Mentions: However, a deeper analysis of Equation (6) shows that it is possible to compute the cumulative row sum for all rows independently and hence simultaneously. This is however not true for Equation (7) due to its dependency on data from the neighboring row. Thus, the best possible system using these equations is to process individual rows in a delayed fashion. As an example, Figure 3 shows a 5 × 5 image for which integral image values are calculated by processing all rows in parallel using these equations. The shaded blocks represent the pixels for which integral image values are computed simultaneously; blocks with a cross sign indicate pixels whose integral image values have already been calculated; and white blocks show pixels for which integral image values still need to be calculated. It can easily be seen that the integral image value for the second pixel in the third row cannot be calculated until the integral image value for the second pixel in the second row is calculated. Figure 4 shows the time delay involved in computation of integral image values for different rows. |
View Article: PubMed Central - PubMed
Affiliation: School of Computer Science and Electronic Engineering, University of Essex, Colchester CO4 3SQ, UK. sehsan@essex.ac.uk.
No MeSH data available.