为了解决当拍摄图像的环境光照较低或较高时,导致的图像对比度较低,给特征点的检测和匹配带来极大的困难这一问题,提出一种基于最大图像熵Gamma校正估计的图像特征点检测和匹配方法。通过预处理算法来增强图像的对比度,将其应用于图像特征点的检测和匹配。在预处理阶段,首先采用对数函数对图像进行归一化,根据设定的阈值将图像分为明亮和黑暗分量;然后分别对两个分量自适应地选择不同的参数进行Gamma校正,并且确定使其熵最大化的校正参数为每个分量的最佳参数;最后将上述参数应用于Gamma校正生成对应原始图像明亮和黑暗区域的矫正图,并进行融合生成增强的图像。对预处理完的图像进行特征点检测和匹配实验。研究结果表明:所提出的算法相较于未处理算法、HE和自适应Gamma算法,特征点检测后的匹配数量在欠曝光图像上分别提升85.7%、26.4%和15.2%,在过曝光图像上分别提升59.4%、12.2%和103.8%。匹配效果也有较明显提升。
Abstract
When the ambient lighting of the captured image is low or high, it will result in low contrast of the image, which will bring great difficulties in feature point detection and matching. To solve this problem, An image feature point detection and matching method based on maximum image entropy Gamma correction estimation is proposed. The contrast of the image is enhanced by a preprocessing algorithm, which is applied to the detection and matching of image feature points. In the preprocessing stage, the image is first normalized using a logarithmic function, and divided into bright and dark components according to a set threshold; then different parameters are adaptively selected for Gamma correction for the two components respectively, and the correction parameter that maximizes its entropy is determined to be the optimal parameter for each component; finally, the above parameters are applied to Gamma correction to generate a corrected map that corresponds to the original image bright and dark regions, and fusion is performed. Finally, the above parameters are applied to the Gamma correction to generate the corrected maps corresponding to the bright and dark regions of the original image, and fused to generate the enhanced image. The pre-processed image is subjected to feature point detection and matching experiments. The results show that compared with the unprocessed algorithm, HE algorithm and adaptive Gamma algorithm, the number of matches after feature point detection is increased by 85.7%, 26.4% and 15.2% on dark images, and 59.4%, 12.2% and 103.8% on bright images, respectively, and the matching effect is significantly improved.
关键词
机器视觉 /
特征点 /
图像熵 /
自适应;Gamma校正;匹配效果
Key words
machine vision /
feature point /
image entropy /
adaptive /
Gamma correction /
matching effect
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
［1］LIU L, XIAO J, WANG Y, et al. A Novel Rock-Mass Point Cloud Registration Method Based on Feature Line Extraction and Feature Point Matching［J］. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 1-17.
［2］舒子超, 曹松晓, 蒋庆, 等. 基于3D视觉的微通道换热器定位与尺寸测量［J］. 计量学报, 2022, 43(9): 1128-1134.
SHU Z C, CAO S X, JIANG Q, et al. Positioning and dimensional measurement of microchannel heat exchangers based on 3D vision［J］. Acta Metrologica Sinica, 2022, 43(9): 1128-1134.
［3］刘鑫行, 王思贤, 孙斌, 等. 基于静力衡量法和光学法的超微容量测量方法对比分析［J］. 计量学报, 2022, 43(10): 1250-1255.
LIU X X, WANG S X, SUN B, et al. Comparative analysis of ultrafine volume measurement methods based on static measurement method and optical method［J］. Acta Metrologica Sinica, 2022, 43(10): 1250-1255.
［4］IMAN A K, LUKE F, GERARD D, et al. A survey of state-of-the-art on visual SLAM［J］. Expert Systems with Applications, 2022, 205: 1-17.
［5］李鹤喜, 李威龙. 基于可变形卷积的双目视觉三维重建［J］. 计量学报, 2022, 43(6): 737-745.
LI H X, LI W L. Binocular visual three-dimensional reconstruction based on deformable convolution［J］. Acta Metrologica Sinica, 2022, 43(6): 737-745.
［6］LIU J, GU K, LI M Z, et al. Calibration Method for Depth Measurement of Nano microstructure in Scanning Probe Microscopy［J］. Acta Metrologica Sinica, 2019, 40(4): 549-556.
［7］HUANG Z, SUN L, LI S, et al. EventPoint: Self-Supervised Interest Point Detection and Description for Event-based Camera［C］//IEEE/CVF Winter Conference on Applications of Computer Vision (WACV). Waikoloa HI, USA,2023.
［8］XIAO Y F, GUO Z X, VEELAERT P, et al. General Image Fusion for an Arbitrary Number of Inputs Using［J］. Sensors, 2022, 22(7): 1-22.
［9］ZHANG Y, DI X, ZHANG B, et al. Better Than Reference in Low-Light Image Enhancement: Conditional Re-Enhancement Network［J］. IEEE Transactions on Image Processing, 2022, 31: 759-772.
［10］陈剑, 孙太华, 黄凯旋, 等. 基于直方图均衡化和卷积神经网络的轴承故障诊断方法［J］. 计量学报, 2022, 43(7): 907-912.
CHEN J, SUN T H, HUANG K X, et al. A bearing fault diagnosis method based on histogram equalization and convolutional neural networks［J］. Acta Metrologica Sinica, 2022, 43(7): 907-912.
［11］GUISIK K, JUNNSEOK K. Deep Illumination-Aware Dehazing With Low-Light and Detail Enhancement［J］. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(3): 2494-2508.
［12］FAYADH A, AMMAR A, ABDULLAH G, et al. Reverse gamma correction based GARCH model for underwater image dehazing and detail exposure［J］. Expert Systems with Applications, 2023, 232: 1-14.
［13］田柯, 马小晶, 贺航. 全局双伽马校正与改进SSA的低光照图像增强方法［J］. 电子测量技术, 2023, 46(9): 124-133.
TIAN K, MA X J, HE H. Global Dual Gamma Correction and Improved SSA for Low Light Image Enhancement［J］. Electronic Measurement Techniques, 2023, 46(9): 124-133.
［14］LI F, HUANG X, CHEN Y, et al. Quantitative algorithm for airborne gamma spectrum of large sample based on improved shuffled frog leaping-particle swarm optimization convolutional neural network［J］. Nuclear Science and Techniques, 2023, 34(7): 146-156.
［15］YUN G Y, LAI W, FAN L, et al. Multislice Time-Frequency image Entropy as a feature for railway wheel fault diagnosis［J］. Measurement, 2023, 216: 1-16.
［16］蔡美芳, 万里勇. 基于自适应Gamma校正的红外图像增强算法［J］. 光学技术, 2022, 48(4): 486-491.
CAI M F, WAN L Y. Infrared Image Enhancement Algorithm Based on Adaptive Gamma Correction［J］. Optical Technique, 2022, 48(4):486-491.
［17］MUHAMMAD T R, DAMING S, HUFSA K. A comprehensive experiment-based review of low-light image enhancement methods and benchmarking low-light image quality assessment［J］. Signal Processing, 2023, 204: 1-12
［18］ZHANG Y C, LI T, LI Q, et al. Image motion deblurring via attention generative adversarial network［J］. Computers & Graphics, 2023, 111: 122-132.
［19］INHO J, CHUI L. An optimization-based approach to gamma correction parameter estimation for low-light image enhancement［J］. Multimedia Tools and Applications, 2021, 80(12): 18027-18042.
［20］周非, 陈帅, 吴凯, 等. 快速跟踪分割辅助的动态SLAM［J］. 仪器仪表学报, 2023, 44(5): 313-321.
ZHOU F, CHEN S, WU K, et al. Dynamic SLAM for Fast Track Segmentation Assist［J］. Chinese Journal of Scientific Instrument, 2023, 44(5): 313-321.
［21］韩硕, 陈晓荣, 张彩霞, 等. 基于改进模板匹配算法的物料计数方法研究［J］. 计量学报, 2022, 43(7): 863-868.
HAN S, CHEN X R, ZHANG C X, et al. Research on Material Counting Method Based on Improved Template Matching Algorithm［J］. Acta Metrologica Sinica, 2022, 43(7): 863-868.
基金
国家自然科学基金联合基金重点支持项目(U21A20486); 中国科学院自动化研究所复杂系统管理与控制国家重点实验室开放课题(20220102); 中央高校基本科研业务费专项资金(2023JC006)
PDF(668 KB)
