• Acta Optica Sinica
  • Vol. 37, Issue 10, 1010001 (2017)
Weili Ding1, Mingkui Wang1、*, Zhao Gu1, and Wenfeng Wang2
Author Affiliations
  • 1 1School of Electrical Engineering, Yanshan University, Qinhuangdao, Hebei 066004, China
  • 2 2School of Vehicles and Energy, Yanshan University, Qinhuangdao, Hebei 066004, China
  • show less
    DOI: 10.3788/AOS201737.1010001 Cite this Article Set citation alerts
    Weili Ding, Mingkui Wang, Zhao Gu, Wenfeng Wang. A Fast Image Correction Method for Multi-Target QR Codes[J]. Acta Optica Sinica, 2017, 37(10): 1010001 Copy Citation Text show less
    Symbolic characteristics of QR codes. (a) Structure of QR codes; (b) detection patterns
    Fig. 1. Symbolic characteristics of QR codes. (a) Structure of QR codes; (b) detection patterns
    Results comparison of QR codes binarization. (a) Original images of QR codes; (b) Otsu algorithm; (c) the proposed algorithm
    Fig. 2. Results comparison of QR codes binarization. (a) Original images of QR codes; (b) Otsu algorithm; (c) the proposed algorithm
    Search the central point of detection pattern. (a) Line scan; (b) column scan; (c) adjust to the center position; (d) judge the relationship between S1 and S2
    Fig. 3. Search the central point of detection pattern. (a) Line scan; (b) column scan; (c) adjust to the center position; (d) judge the relationship between S1 and S2
    Position relationship of QR codes in the image
    Fig. 4. Position relationship of QR codes in the image
    Search of QR codes vertices. (a) Position analysis of detection pattern points; (b) contour tracing start point; (c) vertices of detection patterns; (d) the fourth vertex
    Fig. 5. Search of QR codes vertices. (a) Position analysis of detection pattern points; (b) contour tracing start point; (c) vertices of detection patterns; (d) the fourth vertex
    Inverse perspective transformation
    Fig. 6. Inverse perspective transformation
    Correction results in different environments. (a) Normal illumination; (b) low light environment; (c) complex background environment
    Fig. 7. Correction results in different environments. (a) Normal illumination; (b) low light environment; (c) complex background environment
    Correction results of fuzzy image. (a) Light fuzzy; (b)(c) fuzzy
    Fig. 8. Correction results of fuzzy image. (a) Light fuzzy; (b)(c) fuzzy
    Analysis of different algorithms. (a) Ref. [5] algorithm; (b) Ref. [13] algorithm; (c) Ref. [8] algorithm
    Fig. 9. Analysis of different algorithms. (a) Ref. [5] algorithm; (b) Ref. [13] algorithm; (c) Ref. [8] algorithm
    AlgorithmAverage execution time /msRecognition rate /%
    Group 1Group 2Group 3Group 1Group 2Group 3
    Ref. [9] algorithm96.87963.77679.216857877
    Proposed algorithm32.49722.32827.164929088
    Table 1. Comparison of computational time and recognition rate of multi QR codes correction algorithms
    AlgorithmAverage execution time /msRecognition rate /%
    Group 1Group 2Group 3Group 1Group 2Group 3
    Ref. [5] algorithm37.62336.87938.264706143
    Ref. [8] algorithm34.42332.66737.213827664
    Ref. [13] algorithm29.41427.01628.864847851
    Proposed algorithm11.21310.46411.625939189
    Table 2. Comparison of calculation time and recognition rate of single QR code correction algorithms
    Weili Ding, Mingkui Wang, Zhao Gu, Wenfeng Wang. A Fast Image Correction Method for Multi-Target QR Codes[J]. Acta Optica Sinica, 2017, 37(10): 1010001
    Download Citation