Binary conversion and slant correction algorithm in preprocessing of license plate recognition 車牌識(shí)別預(yù)處理中的二值化及傾斜矯正算法
Editing floating - point values can result in minor inaccuracies because of decimal - to - binary conversion of fractional components 編輯浮點(diǎn)值時(shí),由于要將小數(shù)部分從十進(jìn)制轉(zhuǎn)換為二進(jìn)制,因此所得的結(jié)果可能存在微小誤差。
To this problem a series of image rectification procedure which utilize geometric rectification theory in vehicle license recognition system are pointed out : we first locate vehicle license and execute the edge detection and binary conversion , then proceed calibration in which we use two calibration scheme looking for control point coordinate , one based on the hough transformation , the other based on image analysis method , and finally realize image correction in geometrical transform and bilinear gray interpolation method 針對(duì)此問(wèn)題本文給出了一套將幾何畸變校正技術(shù)應(yīng)用到汽車牌照自動(dòng)識(shí)別系統(tǒng)中的車牌圖像校正方法:首先提取牌照信息,并對(duì)圖像進(jìn)行邊緣檢測(cè)和二值化的預(yù)處理;然后提出了基于hough變換控制點(diǎn)提取和基于圖像分析法的控制點(diǎn)提取的兩種實(shí)時(shí)標(biāo)定方法;最后通過(guò)空間坐標(biāo)變換和灰度插值完成圖像的校正。