• Laser & Optoelectronics Progress
  • Vol. 56, Issue 5, 051001 (2019)
Chao Xu* and Xueliang Ping
Author Affiliations
  • Jiangsu Key Laboratory of Advanced Food Manufacturing Equipment and Technology, School of Mechanical Engineering, Jiangnan University, Wuxi, Jiangsu 214122, China
  • show less
    DOI: 10.3788/LOP56.051001 Cite this Article Set citation alerts
    Chao Xu, Xueliang Ping. Line Detection Algorithm Based on Improved Random Hough Transformation[J]. Laser & Optoelectronics Progress, 2019, 56(5): 051001 Copy Citation Text show less
    Hough transformation diagram
    Fig. 1. Hough transformation diagram
    Flow chart of algorithmic detection
    Fig. 2. Flow chart of algorithmic detection
    8 neighborhoods of P0
    Fig. 3. 8 neighborhoods of P0
    Linear edge detection based on gradient direction difference. (a) Experimental image; (b) 8 neighborhood edge group; (c) edges profile with straight line feature
    Fig. 4. Linear edge detection based on gradient direction difference. (a) Experimental image; (b) 8 neighborhood edge group; (c) edges profile with straight line feature
    Linear edge detection of experimental image. (a) Standard Hough transform; (b) proposed algorithm
    Fig. 5. Linear edge detection of experimental image. (a) Standard Hough transform; (b) proposed algorithm
    Linear edge detection of building image. (a) Building image; (b) standard Hough transform; (c) proposed algorithm
    Fig. 6. Linear edge detection of building image. (a) Building image; (b) standard Hough transform; (c) proposed algorithm
    Linear edge detection based on gradient direction difference. (a) Road image; (b) edge detecting image; (c) standard Hough transform; (d) proposed algorithm
    Fig. 7. Linear edge detection based on gradient direction difference. (a) Road image; (b) edge detecting image; (c) standard Hough transform; (d) proposed algorithm
    TestSize /(pixel×pixel)Time /s
    Standard HoughtransformProposedalgorithm
    Test 1512×3840.9680.389
    Test 2532×5722.2251.153
    Test 3983×6043.9431.075
    Table 1. Algorithm processing time contrast
    LineEndpoint /pixelLength /pixel
    Line 1(193,158) (338,143)145.77
    Line 2(343,147) (352,255)108.37
    Line 3(351,257) (206,269)145.50
    Line 4(203,268) (192,161)107.56
    Table 2. End point coordinates and line segment length