文件列表:

code (0, 2016-08-05)
code\evaluate.m (834, 2016-07-18)
code\fun (0, 2016-07-18)
code\fun\cal_hist_thresh.m (1192, 2015-12-13)
code\fun\cal_hist_two_thresh.m (1085, 2015-12-13)
code\fun\draw_circle.m (936, 2015-12-13)
code\fun\enhance_image.m (650, 2015-12-13)
code\fun\filter_region.m (401, 2013-08-06)
code\fun\find_center_v3.m (3402, 2015-12-13)
code\fun\find_circles_NIR.m (706, 2016-07-18)
code\fun\find_circles_VW.m (527, 2015-12-13)
code\fun\find_radius.m (1733, 2015-12-13)
code\fun\fit_eyelid.m (3109, 2016-07-18)
code\fun\get_pupil_region.m (1207, 2015-12-13)
code\fun\get_reflection_region.m (339, 2015-12-13)
code\fun\get_rtv_l1_contour.m (238, 2015-12-13)
code\fun\get_sector_region.m (516, 2013-07-15)
code\fun\histtruncate.m (3584, 2015-09-26)
code\fun\hough_parabolic.m (1944, 2015-01-07)
code\fun\mask_lower_region.m (1809, 2015-02-04)
code\fun\mask_upper_region.m (211, 2013-08-06)
code\fun\normalize8.m (4159, 2012-01-26)
code\fun\process_ES_region.m (968, 2015-12-13)
code\fun\quality.m (583, 2015-12-13)
code\fun\remove_reflection.m (289, 2016-07-18)
code\fun\rtv_l1_smooth2.m (4587, 2015-12-13)
code\fun\segment_angle_range.m (390, 2015-02-04)
code\fun\smooth.m (253, 2016-06-21)
code\fun\SSR_enhance.m (915, 2015-12-13)
code\fun\thresh_angle_range.m (323, 2013-08-06)
code\normalize (0, 2016-08-05)
code\normalize\circlecoords.m (1622, 2003-11-30)
code\normalize\normaliseiris.m (5035, 2003-11-30)
code\normalize\normalize_iris_sample.m (1108, 2016-08-05)
code\segment (0, 2016-07-18)
code\segment\main_CASIAv4.m (1390, 2016-07-18)
code\segment\main_FRGC.m (1219, 2016-07-18)
code\segment\main_UBIRISv2.m (1139, 2016-07-18)
code\segment\segment_NIR.m (4862, 2016-07-18)
... ...

The code made available in the attachment is the implementation of the iris segmentation approach described in the following paper: Zijing Zhao and Ajay Kumar, "An accurate iris segmentation framework under relaxed imaging constraints using total variation model", Proc. ICCV 2015, pp. 3828-3836, Santiago, Chile, Dec. 2015. Anyone is permitted to use, distribute and change this program for any non-commercial usage. However each of such usage and/or publication must include above citation of this paper. ================ Program Specifications ================ This matlab code is run under Matlab R2012b+ environment using Windows 7 OS on a PC with 2.0 GHz Intel Dual core processor with 2GB RAM. There are three main programs that can be used to generate iris segmentation results from three different databases. main_UBIRISv2.m, main_CASIAv4.m and main_FRGC.m Above three files also define pre-tuned parameters for each of the respective dataset. The paths defined in these three files should be changed accordingly and/or the database-specific parameters should for replaced when the respective programs are used on other datasets. segment_VW and segment_NIR are the main programs for segmenting the iris region from the visible wavelength and near infrared eye images respectively. These two files do not need any changes for their general/respective usage. In addition, there are several funcions used by above two main programs. These functions are stored in the folder 'fun'. Some of these functions require the use of MATLAB Image Processing Toolbox while some functions included in this folder are also made available via cited references (e.g. SSR enhancement) but included for the convenience. This segmentation method can also be used to accurately segment many other iris image databases. We have therefore provided the segmentation programs for some other popular databases, such as ICE2006, CASIA.v3-Lamp, PolyU Cross-Spectral, etc. These additional program callers are contained in the segment_additional folder, where the parameters for these databases are pre-set. However, since the ground truth masks for these databases are not available, these parameters may not necessarily be the optimal and they are just provided for immediate use and benefit of researchers. ================ Resources for Re-production ================ The following resources are made available for re-producing the experimental results described in the above paper. 1. Coordinates for segmenting the eye region from original face images in CASIA.v4-distance databases are available in the attached folder named as 'eye_coordinates'. 2. Manually segmented ground truth iris masks for FRGC and CASIA.v4-distance databases: http://www.comp.polyu.edu.hk/~csajaykr/myhome/research/paper11.zip (above masks are available since 2012 via reference [43] of T-IP 2012 paper titled 'A Unified Framework for Automated Iris Segmentation Using Distantly Acquired Face Images') 3. The eye images for the experiments corresponding to the above FRGC masks can be accessed (only for research use) via http://www.comp.polyu.edu.hk/~csajaykr/FRGC_eye.rar ================ ===================== ================ Disclaimer: This software is provided on 'as it is' basis and does not include any warranty of any kind. Authors: Zijing Zhao (Jason.zhao@connect.polyu.hk), Ajay Kumar (ajay.kumar@polyu.edu.hk) Department of Computing, The Hong Kong Polytechnic University, Hong Kong

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：