会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 1. 发明专利
    • DROWSY DRIVER DETECTION SYSTEM USING AVERAGE LOCAL BINARY STRUCTURE
    • AU2020102438A4
    • 2020-11-12
    • AU2020102438
    • 2020-09-26
    • ABUSHAM EIMADGOYAL MUKTAKUMAR ADARSHMISHRA BIMAL KUMARNAYYAR ANANDSAINI DINESH KUMAR DR
    • SAINI DINESH KUMARABUSHAM EIMADGOYAL MUKTAMISHRA BIMAL KUMARNAYYAR ANANDKUMAR ADARSH
    • G06K9/00G06T7/00G08B21/06
    • A system and method of detecting drowsiness of a driver eyes by average local binary structure is disclosed. A webcam captures the driver's face and detects eyes based on boosted classifiers. The classifiers calculate a difference between the adjacent rectangular regions, such that the difference calculated is considered as the eye detection. An average local binary structure (ALBS) algorithm creates partitions of the eye image into local regions of at least twenty five pixels with a center point being a target pixel. The algorithm labels the pixels of the local regions by thresholding neighborhood of each pixels.The average local features are extracted from a texture region and transmitted to the classifier for recognition, wherein the pixels in case of diagonal direction, first left side and then right side is filtered and in case of linear direction the checking starts from left, above, right and below sides. The average local binary structure (ALBS) removes illumination and extracts local features from the neighborhood by taking geometrically deformed pixels of the image and the extracted local features are transmitted to alert the driver. AVERAGELOCAL EYE DETECTION BINARY (ALBS) 120 DETECTION ILLUMINATION NORMALIZATION MODULE DETECTING THE DRIVER EYE IMAGE BY AN EYE DETECTION MODULE BASED ON BOOSTED CLASSIFIERS TRANSMITTING THE IMAGE FOR ILLUMINATION NORMALIZATION BY THE AVERAGE LOCAL BINARY STRUCTURE (ALBS) ALGORITHM TO CREATE PARTITIONS OF THE EYE IMAGE INTO LOCAL REGIONS OF AT LEAST TWENTY FIVE PIXELS WITH A CENTER POINT BEING A TARGET PIXEL 220 LABELLING THE PIXELS OF THE LOCAL REGIONS BY THRESHOLDING NEIGHBORHOOD OF EACH PIXELS EXTRACTING LOCAL FEATURES FROM THE NEIGHBORHOOD BY TAKING GEOMETRICALLY DEFORMED PIXELS OF THE IMAGE DETECTING DROWSINESS OF EYES BY COMPARING THE PIXELS WITH HIGH GRAY VALUE WITH A THRESHOLD VALUE BEING PREPROGRAMMED IN A LEARNED EYE MODEL AND TRANSMITTING AN ALARM TO THE DRIVER ON DROWSINESS 250