Digital Resources & Technology Partners
Building RIPM’s Digital Resources
A professional scanning laboratory for print and microform has been set up at the RIPM International Center. All scanning is done to preservation standards. As of September 2024, over 4.1 million pages have been scanned.
RIPM possesses three print capture systems consisting of (i) an i2S CopiBook-Open A2 System with an Eagle 71mp camera, (ii) a CZUR ET24 Pro portable scanner, (iii) Canon EOS-1Ds Mark III and Canon 5D Mark II digital SLR cameras, each with a Canon EF 24-70mm f/2.8L zoom lens—which capture images at a true 400 dpi color or grayscale, along with a specially-designed and RIPM-built 120° book cradle with an operator-controlled motorized glass platen. These systems permit RIPM to photograph original materials with a mind to conservation by strictly limiting the stress placed on the volume by the scanning process.
Microfilms are scanned on a Wicks and Wilson RS 200 Rollfilm scanner. Each film is scanned according to the same preservation standards and at a speed which is comfortable for the operator to monitor any shifts in film or scan quality.
Images are stored on a 120tb RAID array which is accessed via gigabit ethernet. Data is backed up into a cloud storage environment using a dedicated storage and backup server. Copies of the RIPM images and associated data are kept in geographically-distant locations.
Ten workstations are used for image processing. RIPM employs i2S's LIMB Server software for image batch processing. Custom-built Linux applications are used to link database records to full-text images. The resulting XML-structured data is audited before publication. Optical Character Recognition (OCR) is accomplished with both ABBYY and Tessaract software. Citation metadata for RIPM Jazz Periodicals is created and edited using custom capture software developed in conjunction with RIPM's technology partner, NES. Images, OCR, and associated metadata are housed on RIPM’s web servers at Towson University, part of the University of Maryland System. For databases available on EBSCOHost, versions of this data is also delivered to EBSCO Information Services. For the RIPM Index with Full Text translation feature, original text is automatically generated by OCR with minimal human correction. Errors may occur. Certain pages, if scanned from imperfect original copies or microfilm, may be more prone to errors. Please report significant or systematic errors to archive@ripm.org.
RIPM’s Technology Partners
NES (NISC Export Services)
NES’s team of engineers provides high-quality application development and metadata services to publishers and educational institutions all over the world. NES expanded upon RIPM’s models and developed a robust search and retrieval system with statistics gathering, standardized subscriber usage reporting, account management, and support for outside OpenURL-compliant linking. NES also built the technologies underlying the RIPM Preservation Series and RIPM Jazz Periodicals interfaces.
NES also provides data generation and processing services. A dedicated team at NES generates citation data for RIPM Jazz Periodicals and RIPM Preservation Series which is then edited at the International Center in Baltimore.
Web link:
http://www.nes.co.in/
Managed Hosting in Strategic Partnerships and Applied Research (SPAR), Towson University
Towson University, part of the University of Maryland System, provides web hosting and systems development for private, public and not-for-profit organizations. SPAR has provided RIPM with a hosting environment composed of Dell Servers, EMC storage and CISCO networks, all connected . SPAR ensures RIPM’s availability through 24/7/365 monitoring, nightly system backups and system-wide redundancies.
Web link:
http://www.towson.edu/outreach/iss/
i2s / Image Retrieval
When scanning, images may be askew, have black borders around pages, and the color may not be ideal. To correct these problems in an efficient manner, RIPM employs i2s’ LIMB batch-processing image correction software. Operating with large batches of files, LIMB automatically deskews images with a high-quality process without leaving artifacts. Automatic cropping is set to remove black borders; coloration can be adjusted either at the batch- or individual page level. Most importantly, LIMB maintains the original image thereby always allowing the user to return to the original. LIMB also enables metadata creation and with the addition of an OCR engine generates full text documents from scans.
In the United States, i2s is exclusively distributed by Image Retrieval.
Web link:
http://www.i2s-digibook.com/ for i2s;
http://www.iiri.com/ for United States distribution.
Wickes and Wilson, part of The Crowley Company
At first glance it may appear that scanning multiple printed books is more demanding than scanning a single reel of microfilm. However, one reel may contain thousands of images of variable quality, many requiring individual attention on the part of the operator. Therefore, one must balance capture speed with the operator’s ability to view and to correct individual images when necessary. A Wickes and Wilson’s RS 200 Rollfilm scanner is in operation at the Center. It utilizes 12-bit image processing to produce high quality images. Using Wickes and Wilson’s ScanFilm software, RIPM can automatically split pages, perform basic cropping, and visually monitor scanning accuracy.
Web link:
The Crowley Company / Wickes and Wilson
Legacy Partners
Image Access / Digital Library Systems Group (DLSG)
For the initial builds of the RIPM Online Archive / RIPM with Full Text, the Digital Library Systems Group of Image Access developed a customized version of their Opus Digitization Workflow software, enabling RIPM to link the RIPM Online database to the scanned pages. Utilizing XML-based linking, the RIPM-DLSG approach allowed for the highlighting of individual parts of a page while displaying it in its entirety. The reader therefore easily perceives the relationship between the hierarchical structure.
Working with RIPM’s staff, the DLSG engineers first created a manual “zoning” application, importing records from the RIPM database and linking each to a page(s) or part thereof. The link included periodical name, publication date, volume, issue, page(s), and location on the page(s). To introduce automation to the process, DLSG engineers created a semi-automated process which utilizes OCR and layout-recognition technology. Both manual and automated steps are subject to manual review before exporting the XML data and creating web-accessible derivative images.
Web link:
http://www.imageaccess.com/