技術(shù)要點

• 程序開發(fā)環(huán)境：Win32 visua開發(fā)環(huán)境
• 提供的示例的源文件包含：VB.NET, C#, Delphi .NET, VB, VC++
• 在.NET中作為一個可管理的控件來配置（詳細內(nèi)容查看白皮書《在Microsoft .NET平臺中構(gòu)建健壯的圖像組件》）
• 允許在所有支持ActiveX/COM控件的開發(fā)環(huán)境中使用
• 為多種操作提供可靠的反饋值
• 使用可配置的調(diào)試設(shè)置選項，查看控件的運行細節(jié)信息
• 提供了免費的，功能完整的體驗版控件，您可以立即下載體驗。

圖像輸入，圖像輸出和圖像操作

• 通過內(nèi)置的ImagXpress Professional .NET 組件和ActiveX/COM 控件實現(xiàn)。 (詳細內(nèi)容，查看完整的 ImagXpress Professional v7 產(chǎn)品說明)
• 功能特點包括：查看、壓縮、轉(zhuǎn)換、TWAIN掃描、注釋、打印、圖像操作等

抗扭曲

• 自動拉直扭曲的圖像，并反饋糾正的角度
• 圖像的抗扭曲處理只需零點幾秒的時間
• 直接進行圖像內(nèi)容的處理，而非僅僅處理文檔的邊框
• 用戶可以設(shè)置高質(zhì)量或者高速度選項
  
  • 高質(zhì)量設(shè)置項，在圖像中設(shè)置了大量的點來校驗圖像，最終提供了高質(zhì)量的抗扭曲效果，但是對圖像處理速度會有影響
  • 高速度設(shè)置項，在圖像中只設(shè)置了必要數(shù)量的點來校驗圖像，最終提供了高速度的抗扭曲效果，但是對圖像抗扭曲質(zhì)量會有影響

消噪

• 消除圖像中隨機出現(xiàn)的斑點和噪點
• 用戶允許指定需要處理的斑點大小或者使用自動設(shè)置
• 指定針對單個點圖像的清除算法，避免消除必要的點圖像，比如常規(guī)的標(biāo)點符號

清除點偏差

• 自動清除文字周圍的偏差點的痕跡，提高OCR識別的精確度
• 清除在掃描有彩色背景的圖像時產(chǎn)生的噪點（灰色、藍色等）
• 在不破壞文本內(nèi)容的基礎(chǔ)上，減少由于附加象素點導(dǎo)致的文件大小

　
清除孔洞和斑點

• 清除掃描時出現(xiàn)的孔洞和大塊黑色區(qū)域
• 用戶允許設(shè)置處理范圍
• 指定為整個頁面或者您感興趣的某個區(qū)域
• 根據(jù)需求減少文件尺寸保存清理后的頁面
• 返回檢測到的對象總數(shù)

清除直線

• 自動清除由掃描窗口和文件時產(chǎn)生的實線和線段（同時支持橫向和豎向）
• 在窗體進程中自動修復(fù)有直線打斷的文本內(nèi)容（重新填入被破壞的字符）

　
清除組合框（主要用于ICR窗體中）

• 自動清除窗體中的整個組合框數(shù)據(jù)
• 在窗體進程中自動修復(fù)有直線打斷的文本內(nèi)容（重新填入被破壞的字符）

字符獲取和平滑處理

• 修復(fù)、平滑和加粗不完整的字符
• 修復(fù)破損的字符，創(chuàng)建完整的字符
• 針對點陣字符有較好效果
• 平滑粗糙的字符邊緣
• 根據(jù)需要加粗或者細化處理字符

自動反白文本糾正

• 檢測和自動轉(zhuǎn)換反白顯示的文本區(qū)域
• 將“黑底白字”的文本轉(zhuǎn)換為常規(guī)的“白底黑字”的文本
• 識別任意形狀的反白區(qū)域，包括圓形區(qū)域
• 清除文本中填充的顏色
• 減少由不必要的黑色象素導(dǎo)致增加的文件尺寸

自動求反

• 自動報告“白底黑圖”（正常）或者“黑底白圖”（反白）的圖像
• 自動校正（您只需將選項從“自動報告”改為“自動校正”）
• 創(chuàng)建其他ScanFix image 圖像操作中要求的“黑底白圖”的圖像

指定二進制轉(zhuǎn)換

• 自動將灰度或者彩色圖像轉(zhuǎn)化為黑白圖像，并清理圖像噪點
• 二進制轉(zhuǎn)換選項可以由用戶自行設(shè)定

自動邊框裁剪

• 將黑色邊框自動轉(zhuǎn)化為白色
• 減少圖像大小（同時針對物理尺寸和壓縮文件大小）
• 自動邊框裁剪（您只需將選項從“自動報告”改為“自動校正”）
• 可用于縮微拍攝或過掃描的小尺寸的圖像文件

空白頁/空白矩形區(qū)域檢測

• 檢測并報告空白頁或者您感興趣的空白區(qū)域
• 將選項設(shè)置為適應(yīng)滲漏或者噪點圖像（比如忽略孔洞）
• 根據(jù)需求減少文件尺寸保存清理后的頁面

子圖像操作

• 在主要的操作過程中，隔離原始圖像中某一個區(qū)域進行優(yōu)先處理
• 提高處理速度，僅僅輸出您需要查看的圖像區(qū)域

報告操作結(jié)果

• 針對窗體識別操作，報告需要清除的直線、點偏差區(qū)域和反白文本區(qū)域的位置
• 窗體識別操作和注冊中，非常有用的操作解決方案

附加的圖像操作：

• 倒轉(zhuǎn)
• 鏡像
• 選擇
• 平滑縮放
• 按比例縮放
• 分辨率設(shè)置

Technical Notes

• Programming environments: Win32 visual development environments
• Sample code is included for: VB.NET, Visual C#, Delphi .NET, VB, VC++
• Deploys within .NET as a managed control (see "Building Robust Imaging Components for the Microsoft .NET Platform" white paper)
• Can be used in any development environment that hosts ActiveX/COM controls
• Confidence values are available for many operations
• User configurable debug settings to control level of detail
• Free full-featured trial version available for immediate download

Image Input, Image Output, and Image Handling

• Provided through the included ImagXpress Professional .NET component and ActiveX/COM control (Read the Full ImagXpress Professional v7 Product Description)
• Features include image viewing, compression, conversion, TWAIN scanning, annotation, printing, image processing, and more

Deskew

• Automatically straighten crooked images and report correction angle
• Deskew images in a fraction of a second
• Uses content of the images, not just the document
• User configurable High Quality vs. High Speed setting
  • High Quality sets a higher number of places to check around the image, ultimately providing the highest level of deskew accuracy, but the image processing speed is slower
  • High Speed sets a lower number of places to check around the image, ultimately providing the fastest image processing speed, but offers a slightly lower level of accuracy

Noise Removal

• Remove random specks or noise from images
• Configurable user specification of speck size, or use automatic settings
• Special Isolated Despeckle algorithm prevents removal of desired dots such as punctuation

Dot Shading Removal

• Automatically remove dot shading surrounding text for improved OCR accuracy
• Remove scanner noise caused by colored backgrounds (gray, blue, etc.)
• Reduce files sizes caused by additional pixels, without destroying text

Hole Punch & Blob Removal

• Remove hole punches and larger areas of black
• User configurable threshold settings
• Specify full page or a specific area of interest
• Reduce file sizes and storage requirements with cleaner pages
• Return count of objects found

Line Removal

• Automatically remove solid or broken lines (both horizontal and vertical) from scanned forms and documents
• Automatically repair text that intersects with lines (fill broken characters), essential for forms processing

Comb Removal (commonly used in ICR forms)

• Automatically remove data entry combs from forms
• Automatically repair text that intersects with lines (fill broken characters), essential for forms processing

Character Completion & Smoothing

• Repair, smooth and thicken incomplete characters
• Repair broken or incomplete characters to create complete characters
• Especially good on poor dot matrix text
• Smoothes ragged edges on letters
• Thickens or thins letters as needed

Auto Inverse Text Correction

• Detect and automatically convert inverted text zones
• Change white-on-black text to normal black-on-white
• Recognize inverse zones in any shape, including circular
• Remove the fill color of text
• Reduce file sizes by eliminating unnecessary black pixels

Auto Negate

• Automated reporting of black on white (normal) or white on black (negative) images
• Automated correction (user configurable to simply report or to automatically correct)
• Creates black on white images, which are often required for many other ScanFix image processing actions

Auto Binarize

• Automatically convert grayscale and color images to black and white, and clean up noisy images
• Optimize binarization results using the user configurable options

Auto Border Crop

• Automatically convert black borders to white
• Reduce image file size (both physical dimensions and compressed file size)
• Automated border crop (user configurable to simply report or to automatically correct)
• Use on microfilm or small document images that have been overscanned

Blank Page / Blank Rectangle Detection

• Identify and report blank pages, or blank areas of interest
• Configure settings to accommodate bleed-through or noisy images (i.e. ignore hole punches)
• Reduce file sizes and storage requirements by eliminating blank pages

Sub-Image Processing

• Separate out an area from the original document image prior to image processing
• Speeds up processing and outputs only the smaller area of interest

Reporting of Process Results

• Reports the location of lines, shaded regions and inverse text regions for forms recognition
• Useful in forms processing solutions for form identification and registration

Additional Image Processing

• Flip
• Mirror
• Rotate
• Smooth-Zoom
• Scale
• Set Resolution