Scan, upload, preserve: Mobile app to decentralise digitisation of Assamese language books

# App Digitises Assamese Books to Save Heritage

On April 29, 2026, technology developers and linguistic researchers launched a groundbreaking mobile application designed to decentralise the digitisation of Assamese language books. Addressing the urgent need to preserve decaying physical texts, the platform allows citizens across Assam to scan, upload, and digitally archive literature using their smartphone cameras. By shifting from institutional scanning to a community-driven crowdsourcing model, the initiative aims to rescue thousands of rare books from environmental degradation. Powered by advanced optical character recognition (OCR) specifically trained for the Assamese script, this tool promises to transform physical pages into a permanent, searchable digital library accessible globally. [Source: Hindustan Times].

## The Race Against Time for Assam’s Literary Heritage

The northeastern Indian state of Assam boasts a vibrant literary history that dates back centuries, from the philosophical manuscripts of Srimanta Sankardev to the pioneering 19th-century publications like *Orunodoi*, the first Assamese-language magazine. However, preserving this heritage has historically been an uphill battle. The region’s subtropical climate—characterised by heavy monsoons, high humidity, and prevalent fungal and insect activity—serves as a natural enemy to paper and traditional *Sanchi Pat* (tree bark) manuscripts.

For decades, state libraries, academic institutions, and private collectors have struggled with the immense physical and financial burden of preservation. Conventional digitisation efforts require expensive overhead scanners, controlled lighting, and dedicated personnel. Consequently, thousands of out-of-print books, early 20th-century journals, and local histories have remained locked in private cupboards or dusty library shelves, steadily deteriorating.

“We are losing critical fragments of our cultural memory every single monsoon,” notes Dr. Arindam Barua, an independent archivist and historian. “Institutional digitisation is necessary but slow. If we do not accelerate the process, an entire century of regional thought, folklore, and socio-political commentary will quite literally turn to dust.”

## How the Decentralised Platform Works

The newly launched mobile app operates on a simple, user-friendly premise: “Scan, upload, preserve.” Recognising that smartphone penetration in rural and semi-urban Assam has reached unprecedented levels, the developers have placed a powerful archiving tool directly into the hands of the public.

When a user identifies a rare or out-of-print Assamese book, they can use the app to photograph the pages sequentially. The software features an intelligent auto-cropping and deskewing camera interface that guides users to capture flat, high-contrast images. Once the scanning is complete, the images are compressed to save mobile data and uploaded to a secure, cloud-based repository.

**Key Features of the Mobile Application:**
* **Edge Detection & Auto-Capture:** Ensures pages are captured without warped borders or background clutter.
* **Offline Mode:** Users in remote areas with poor internet connectivity can scan books offline and sync the data once they reach a stable network.
* **Metadata Tagging:** Contributors can add the author, publication year, genre, and physical condition of the original text.
* **Community Moderation:** Uploads undergo a peer-review process where volunteers verify image clarity and copyright status before public release.

## Breaking the Optical Character Recognition Barrier

The core technological breakthrough of this initiative lies in its backend artificial intelligence. Capturing an image of a page is only half the battle; to make the archive truly useful for researchers, the text must be searchable. This requires Optical Character Recognition (OCR), a technology that historically struggled with Indic scripts.

The Assamese alphabet, while sharing its roots and visual similarities with Bengali, possesses unique characters such as ‘ৰ’ (ra) and ‘ৱ’ (wa), alongside hundreds of complex conjunct consonants (*juktakkhor*). Older books often feature degraded typography, faded ink, and outdated spelling conventions, confusing standard OCR engines.

“Prior OCR models would output gibberish when faced with 1940s Assamese print,” explains Manash Saikia, a lead AI researcher on the project. “We trained our neural networks on thousands of manually transcribed pages, teaching the AI to understand the context of the language. Our current model features a 96% accuracy rate, even on yellowed, brittle pages. It automatically converts the image into Unicode text.” [Additional: Public records on Indic language AI processing].

## Empowering Citizens as Digital Archivists

By decentralising the digitisation process, the app taps into a deep sense of cultural pride among the Assamese public. The initiative essentially transforms everyday citizens into active custodians of their heritage. College students, retired teachers, and local literature enthusiasts in districts like Majuli, Dibrugarh, and Barpeta are forming micro-communities to track down endangered books in their respective localities.

This gamified and community-driven approach drastically cuts down logistical costs. Rather than shipping fragile books to a central facility in Guwahati—risking further damage during transit—the preservation happens at the source. This crowdsourced effort echoes the philosophy behind Wikipedia, relying on collective effort to build an open-access repository of human knowledge.

Furthermore, it bridges a significant digital divide. While English and Hindi have vast, searchable digital corpora that fuel everything from academic research to large language models (LLMs), regional languages under the Eighth Schedule of the Indian Constitution often lack a sufficient digital footprint. This initiative actively expands the digital presence of the Assamese language on the internet.

## Navigating Copyright and Public Domain Challenges

A critical component of this decentralised archiving involves navigating the complex web of intellectual property rights. The platform enforces strict guidelines regarding the types of books that can be fully digitised and made publicly available.

| Copyright Status | App Action / Availability |
| :— | :— |
| **Public Domain** (Author deceased > 60 years) | Full text scanned, converted to Unicode, and available for free public download. |
| **Out-of-Print / Orphan Works** | Scanned for preservation. Metadata available, but text access may be restricted to verified researchers until clearance. |
| **Active Copyright** | Users are prevented from uploading recent commercial works to protect publishers and authors. |

To handle these nuances, the app features a built-in database of registered Assamese literature. If a user attempts to scan a book published in the last few decades, the app will flag the title, directing the user instead toward finding older, public domain works.

## Global Implications for Education and Research

The impact of a decentralised digital library extends far beyond the borders of Assam. For the global Assamese diaspora living in the United States, Europe, and Australia, accessing original literature has long been a challenge. Parents wishing to introduce their children to indigenous folk tales (like *Burhi Aair Sadhu*) or classical literature can now access these materials instantaneously from the cloud.

Moreover, the creation of a massive, accurate Unicode corpus for Assamese has profound implications for modern computing. Searchable digital text is the foundational data required to train more sophisticated machine translation tools. As this crowdsourced archive grows, platforms like Google Translate and India’s indigenous *Bhashini* project will have access to richer, more contextually accurate linguistic data, further integrating Assamese into global digital ecosystems.

Linguists and historians also stand to gain immensely. Textual analysis that previously took years of manual reading can now be accomplished through basic keyword searches. Scholars studying the linguistic evolution of Assam during the early 20th century can seamlessly track how specific words, political ideas, and cultural motifs spread through regional literature.

## Conclusion: A Blueprint for Linguistic Preservation

The launch of the Assamese digitisation mobile app marks a watershed moment in cultural preservation. By marrying advanced AI character recognition with a decentralised, community-driven workflow, this tool provides a sustainable lifeline for an endangered body of literature.

More importantly, it offers a replicable blueprint for thousands of other regional and indigenous languages worldwide that face similar threats of physical decay and digital marginalisation. As the project scales, developers aim to refine the technology to capture older, handwritten manuscripts, ensuring that the voices of the past remain legible for the generations of the future. The success of this app demonstrates that the responsibility of preserving history no longer rests solely on underfunded institutions—it now rests in the pockets of the people.

***

By Senior Correspondent, Heritage Desk, April 29, 2026