Overview
Top  Next

Recogniform Image Processor is the complete solution for batch image processing, allowing to execute image enhancement and image transformation on gropus of monochrome, grayscale or color images.

Recogniform Image Processor inherits the effectiveness and the elaboration power of our SDK image processing libraries and moreover user doesn't need to create a new application every time it need a new processing pipe: simply using a custom script the goal is reached.

It's very easy to use the Recogniform Image Processor visual interface for creating the images batch and for defining and customizing processing parameters.

You can create a set of images, the batch, simply dragging files from your folders and dropping them in input images list, or using the built-in dialogs allowing to select files and directories. Also you can import .txt files containing lists of image files to process.

The image processing is script-oriented and this means that you can write the sequence of image processing function to apply on your images. The scripting language used is pascal-like: simple but very powerful. For usual image processing task you can avoid to write manually lines of code and using the visual interface you can leave the task to the application: just select from popup menu the image processing function you want to use, set the parameters required to fine-tune the operation in the next dialog box, and you will see processing script instantly self-generated !
Applying the righ image processing on your images you will obtain great benefits: Optical Character Recognition accuracy is improved, file size decreases, content appearence is better: if you have to make form processing, data capure, OCR, ICR or simply store your images you need this tool !


Features List

·Batch processing  
Simply create your images batches and start the processing ! It's ideal solution to process high volumes of images: the software can run unattended overnight. You can insert specific files in a batch, or full directories, or import pre-built list of files.  
·Multi-threading  
Maximize the throughput with multi-thread technology. You can choice up to eight concurrent "agents" to use in the batch.  
·Comprehnsive file formats support  
All standard file formats are supported: TIFF (Uncompressed, Jpeg Compressed, Packbits, CCITT G3, CCITT G4), JPEG (JFIF), BMP, PNG, PDF (CCITT G4 and JPEG).  
·Full bit depth range  
Bitonal, grayscale and color images are supported. The color depth can be 1, 4, 8 or 24 bits per pixel !  
·Powerful scripting language  
A pascal-like language easy to learn and to use, with built-in wizard to generate automatically scripts. You can use variables, library functions, conditional and loop statements, and more !  
·Large functions library  
Hundreds of functions ready to be used in your processing scripts: image processing and management, data handling and conversion, files and database access, DDE, system and more...  
·Conditional processing  
Using if-then-else scriping construct you can decide in run-time to make some processing operation only and when some conditions are verified.  
·Zonal processing  
You can define your region of interest: extract a sub-image, process only this area, output processed zone and/or re-applying changes on unmodified original.  
·Customizable log  
You can decide the informations to insert in the log, reporting the process results you need.  
·Dual Image Viewer companion  
At no extra charge is enclosed a viewer utility allowing to see each image before and after the processing: it's ideal to save time making an interactive quality control ! You can zoom and pan the two version of the same image at the same time with one single mouse click.  

Image Processing Functions List
 
·Deskew - Straighten crooked images! Using high-volume scanners, automatic document feeder inevitably skew paper: you can fix the problem and obtain straighten images without rescan, correcting skew in full automatic way at high speed. Up to 45 degrees of skew are handled and the angle can be estimated with great accuracy in two ways: analyzing the text inside the image or finding the black border around the paper. To get more info on deskew you can give a look to Deskew SDK  
·Despeckle and Noise Removal - Removes speckle cleaning images! Scanning paper copies and microfilms, dot shading, or just dust and dirty can generate noisy images: you can fix the problem with automatic noise removal.You can determinate how large an area of speckle can be, fine-tuning the process for your images.To get more info on despeckle you can give a look to Despeckle SDK  
·Black Border Removal and Auto Cropping - Remove black border around images! Using microfilm scanner or paper scanner with black background, this black border is produced around the image: you can fix the problem and obtain cleaned images without rescan. You can select a threshold level and choice to remove the dark border cropping the image or simply turn it white.  
·Deformation Correction - Correct deformated images! Using microfilm or microfiches scanners is usual to get deformed images, looking as parallelogram shape instead then a rectangular shape. You can fix the problem and obtain geometrically correct images.  
·Inverse Text Correction - Change white-on-black text into usual black-on-white text ! You can fix inverted text zone allowing OCR systems to read all the data without skipping this dark boxes. A function to invert full image is also available.  
·Line Removal - Remove lines, horizontal, vertical or both! When scanning forms is very important to remove pre-printed lines before making OCR/ICR to get an acceptable recognition rate. You can set several parameters to remove only the elements you need to remove: minimum line lenght, minimum length/thickness ratio, max number of interrutions. The system is so smart that characters crossed by a line are repaired and reconnected after the remotion.  
·Streak removal - Remove streaks ! You can eliminate very thin vertical or horizontal dark or white wires generated during the scan.  
·Smoothing - Repairs and complete broken or incomplete characters and lines! You can fix ragged images featuring poor text/drawings quality without reascan, obtaining smoothed images.  
·Erosion and Dilation - Erode or dilate images ! Make readable dot matrix characters using dilation filter or disconnect dark characters using erosion filter.  
·Light Balancing -Balance the light ! Increse or decrease contrast and brightness by a custom percentage or ask the system to equalize this values automatically. Adjust gamma curve to better color/grayscale visualization.  
·Registration -Register images, repositioning the content in the page! When scanning forms is very important to register image before to extract fields and zones for processing. You can shift the image horizontally and vertically as you need.  
·Resizing and Scaling - Resize or scale images! You can set the width and height of the output image and choice if preserve the aspect ratio. Also you can resize the image but not the content or you can scale image and content. In this last case you can enable interpolation to get better result. A specific function to get grayscale thumbonails from monochrome images is also available.  
·Rotation and Flipping- Rotate images ! You can rotate images by any angle.You can flip images horizontally or vertically.  
·Conversion -Convert images ! Convert images between true-color, grayscale or bitonal classes with great quality.  
·Fixed Thresholding, Automatic Thresholding, Dynamic Thresholding, Edge Thresholding - Threshold graylevel images producing monochromatic images! You can choice the preferred method, set custom parameters and obtain perfect results. To get more info on thresholding you can give a look to Thresholding SDK  
·Dithering - Dithering your grayscale images! Convert your gray image in monochrome preserving halftones.  
·Convolutin and Filtering - Convolve the images using a 3x3 customizable filter! Apply image filter using a 3x3 matrix. You can find enges, make image sharpen, and more...  
·Color Dropping - Drop a specific color! Instead than rescan using an Ink Drop-Out Lamp, use your color images to get a grayscale or a bitonal image with the colored form removed. You can select the color to remove (Red, Green, Blue, Cyan, etc...) or find in runtime the dominant color to remove.  
·Quality Control - Make automatic quality control! Evaluate in runtime image properties as brightness, contrast, variance, dominant color, size, bit dept and skew.