• Laser & Optoelectronics Progress
  • Vol. 60, Issue 12, 1210022 (2023)
Shuangshuang Zheng, Wenxue Wei*, and Cong Xu
Author Affiliations
  • School of Computer Science and Engineering, Shandong University of Science and Technology, Qingdao 266590, Shandong, China
  • show less
    DOI: 10.3788/LOP221707 Cite this Article Set citation alerts
    Shuangshuang Zheng, Wenxue Wei, Cong Xu. Low Illumination Image Enhancement Algorithm Combining Total Variation and Gamma[J]. Laser & Optoelectronics Progress, 2023, 60(12): 1210022 Copy Citation Text show less
    Flow chart of proposed algorithm
    Fig. 1. Flow chart of proposed algorithm
    Comparison of reflection images constrained by different TV models. (a) Original drawing; (b) original TV model; (c) proposed TV(r=3); (d) proposed TV (r=0.3)
    Fig. 2. Comparison of reflection images constrained by different TV models. (a) Original drawing; (b) original TV model; (c) proposed TV(r=3); (d) proposed TV (r=0.3)
    Comparison of illumination correction images.(a) Original drawing; (b) μ=0.5; (c) adaptive μ value
    Fig. 3. Comparison of illumination correction images.(a) Original drawing; (b) μ=0.5; (c) adaptive μ value
    Comparison of Retinex decomposition. (a) Original images; (b) model of reference [15]; (c) STAR; (d) WVM; (e) SSR; (f) proposed model
    Fig. 4. Comparison of Retinex decomposition. (a) Original images; (b) model of reference [15]; (c) STAR; (d) WVM; (e) SSR; (f) proposed model
    Contrast of enhanced images. (a) Original images; (b) algorithm of reference [11]; (c) algorithm of reference [15]; (d) STAR; (e) WVM; (f) MSRCR; (g) algorithm of reference [13]; (h) proposed algorithm
    Fig. 5. Contrast of enhanced images. (a) Original images; (b) algorithm of reference [11]; (c) algorithm of reference [15]; (d) STAR; (e) WVM; (f) MSRCR; (g) algorithm of reference [13]; (h) proposed algorithm
    Comparison of safety helmet and human body detection between low illumination image and enhanced image. (a) Original image detection and recognition results; (b) enhanced image detection and recognition results
    Fig. 6. Comparison of safety helmet and human body detection between low illumination image and enhanced image. (a) Original image detection and recognition results; (b) enhanced image detection and recognition results
    NIQEImage 1Image 2Image 3Image 4Average
    Model of reference[119.79310.72110.60610.16510.321
    Model of reference[1510.90611.35711.54110.34111.036
    STAR9.2889.94611.9709.07510.070
    WVM11.63210.01810.09611.49010.809
    MSRCR10.23211.45410.64013.26311.397
    Model of reference[1310.87710.1799.6659.89210.153
    Proposed model9.08010.0479.4588.2509.209
    Table 1. Quantitative comparison of NIQE
    SSIMImage 1Image 2Image 3Image 4Average
    Model of reference[110.00370.03940.0020.01700.0155
    Model of reference[150.70320.55510.83410.70280.6988
    STAR0.80100.60500.91050.80690.7809
    WVM0.60470.42500.76440.67500.6173
    MSRCR0.41590.70100.58900.69630.6010
    Model of reference[130.78870.74470.85280.86620.8131
    Proposed model0.82130.79390.92640.87610.8544
    Table 2. Quantitative comparison of SSIM
    PSNRImage 1Image 2Image 3Image 4Average
    Model of reference[115.2038.1168.1537.9927.366
    Model of reference[1514.41113.44517.05514.31414.806
    STAR16.76814.25320.57916.16216.941
    WVM12.07311.67815.39813.45913.152
    MSRCR16.05018.73715.50615.00116.324
    Model of reference[1310.87710.1799.2959.89210.061
    Proposed model20.53419.15520.65622.89920.811
    Table 3. Quantitative comparison of PSNR
    Shuangshuang Zheng, Wenxue Wei, Cong Xu. Low Illumination Image Enhancement Algorithm Combining Total Variation and Gamma[J]. Laser & Optoelectronics Progress, 2023, 60(12): 1210022
    Download Citation