scispace - formally typeset
Search or ask a question
Proceedings ArticleDOI

Memory Efficient ASIC Implementation of Line Hough Transform

01 May 2018-
TL;DR: A MATLAB model of LHT with variable word length is developed, which uses line detection by line Hough transform, which is most efficient technique, which give clear output for a given edge detected image even when there is an appreciable noise present.
Abstract: The innovations and applications in automation, robotics and computer vision are enhanced by the image processing techniques such as object detection, face recognition, image compression etc. This work uses line detection by line Hough transform (LHT). LHT is most efficient technique, which give clear output for a given edge detected image even when there is an appreciable noise present. Normally the LHT requires a large memory for rho and theta calculations. Therefore to reduce the memory requirements rho and theta calculations are implemented with CORDIC Algorithm based functional units. A MATLAB model of LHT with variable word length is developed. The number of bits used to represent the pixel values in edge detected image are varied from 1o Bits to 2o Bits; to trade off the quality of edge detection and time required. This simulation reveals that time for 1o bits is 18.1o29s, 15 bits is 18.2982s and for 2o bits is 19.22s. Here the minimum word length of 1o bits is considered and the same architecture is implemented in TSMC 9onm technology. This requires 4174 MB memory and 5ms as computation time compared to 8384 MB and 2ms. So from the CORDIC based LHT uses lesser memory and relatively longer computation time than the previous implementations
References
More filters
Journal ArticleDOI
TL;DR: A novel curvature aided Hough transform for circle detection (CACD) algorithm, which estimates the circle radius from curvature, which is more practical and less time consuming.
Abstract: Curvature radius is adopted to improve the H-transform for circle detection.Curvature pre-estimation avoids senseless accumulation operation, work faster.The CACD is capable to detect circles of different radius in complex scene."Statistic deviation" is defined to measure the saliency of circle center. Conventional Hough based circle detection methods are robust, but for computers in last century, it is to slow and memory demanding. With the rapid development of computer hardware, Hough transform is acceptable now. Improvement on Hough based circle detection is valuable. In this paper, we present a novel curvature aided Hough transform for circle detection (CACD) algorithm, which estimates the circle radius from curvature. Curvature pre-estimation is capable to avoid both accumulating operations of all the points and interruption between different scales, which result in faster and more precise circle detection. Compared to the conventional Hough-based algorithm for circle detection, the algorithm is more practical and less time consuming. Its time taking is about 1/8 of that of conventional algorithm. Test results on traffic sign images shown that The CACD gets an AUC (Area Under Curve) of 0.9125. The CACD is capable to detect circles of different radius in complex scene.

63 citations

Journal ArticleDOI
TL;DR: The induced architecture presents a high degree of regularity, making its VLSI implementation very straightforward, and may be achieved by generator program, assuring a shorter design cycle and a lower cost.

49 citations

Proceedings ArticleDOI
14 Jul 2012
TL;DR: Analysis shows that this new concentric circle detection algorithm reduces time complexity and improves anti-interference compared with the traditional concentriccircle detection algorithm based on chord mid-point Hough transforms.
Abstract: A new approach of concentric circle detection is proposed in this paper. Firstly, the image is preprocessed by denosing, edge detection, and then the circle centers are allocated by the gradient Hough transform, at last, the radius are detected by the improved one-dimensional Hough transform. The detection efficiency is enhanced by image discretization and reduced resolution ratio in the process of circle center detection, and proves that the circle center is on the gradient line of circle edge points; meanwhile, the radius detection accuracy is improved by merging the similar radius in the process of radius detection. Experimental results show that the method combined with gradient Hough transform and one-dimensional Hough transform has good reliability and is high adaptive to noise, distortion, area of incomplete, edge discontinuous. Analysis shows that this new concentric circle detection algorithm reduces time complexity and improves anti-interference compared with the traditional concentric circle detection algorithm based on chord mid-point Hough transforms.

41 citations


"Memory Efficient ASIC Implementatio..." refers background in this paper

  • ...Last stage b0th the c0mp0nents are added t0 calculate the parameter r which is represent a line.[2,3,4]...

    [...]

  • ...Size 0f mem0ry depends 0n the sampling 0f r parameter.[2,3,4]...

    [...]

Proceedings ArticleDOI
26 Oct 1997
TL;DR: Line extraction by the Hough transform of 16-K point Hough space for a 256/spl times/256 picture can be carried out at video-rates by using a Hough board with a single CAM chip.
Abstract: We describe real-time straight line extraction system and evaluate it. The main part of our system involves a highly parallel processing board on which a highly parallel Hough transform algorithm is mounted. We used the Hough transform, because it useful in line detection. In this board, a content addressable memory (CAM) LSI with dedicated functions for highly parallel image processing is used in this board. Both voting and peak extraction, which make up the Hough transform, are directly executed by the CAM. The CAM acts as a PE (processing element) array that performs highly parallel processing for the Hough transform and it also acts as a memory for two-dimensional Hough space. Sixteen thousand Hough points can be mapped at once and thousands more can be mapped by time-sharing in one CAM. Also variable Hough point resolution is available, because the CAM-based HT algorithm on this board is scalable. Thus line extraction by the Hough transform of 16-K point Hough space for a 256/spl times/256 picture can be carried out at video-rates by using a Hough board with a single CAM chip. Moreover a Hough transform with weighted voting for noisy images can also be achieved at video-rates.

27 citations

Proceedings ArticleDOI
22 Oct 2012
TL;DR: The proposed architecture enables storing the HT space on the FPGA's memory blocks with no need for accessing external memory while processing large size images in real-time with high frame rate.
Abstract: Hough transform (HT) is a widely used algorithm in machine vision systems. In this paper, a memory efficient architecture for implementing HT on FPGAs is presented. The proposed architecture enables storing the HT space on the FPGA's memory blocks with no need for accessing external memory while processing large size images in real-time with high frame rate. It can be used for both line and circle detection. Results show very good accuracy with images processed at 30 fps frame rate and image size of 800 × 600. This compares favourably with other reported architectures in the literature.

25 citations


"Memory Efficient ASIC Implementatio..." refers background or methods in this paper

  • ...Last stage b0th the c0mp0nents are added t0 calculate the parameter r which is represent a line.[2,3,4]...

    [...]

  • ...After getting the sine and cosine values through cordic algorithm we can use LHT....

    [...]

  • ...Size 0f mem0ry depends 0n the sampling 0f r parameter.[2,3,4]...

    [...]

  • ...S0 we need an0ther scan 0f image f0r finding them after applying LHT.[3,4] The array size we need is R*N and let the image size be 64o*48o pixels....

    [...]

  • ...ELS requires more c0mputation time but d0es n0t require large mem0ry space.[3,4]....

    [...]