Regularly, we receive questions about how our text recognition service can be implemented in existing software. Sometimes it can be an extensive puzzle to determine in which stream and which format the Klippa OCR service should be used. In this blog I take you through some steps to determine the best flow.
Current software architecture
The Klippa OCR software is always an integration into your existing system. We always have to deal with API connections or webservices. Over the years, we’ve become a real specialist in that field. Our OCR API can be used in combination with all common programming languages such as PHP, Java and .NET. The OCR service is made available through a RESTful API. Every web-based client can integrate with our API.
OCR Input format and input current
First of all, it is good to determine where you want to place the OCR software in your current flow. In some cases, a photo is made in-app, which needs to be recognized. [If you’re looking for a solution for that, please check our OCR SDK page 😉 ]. In that case, you can send the image directly to us, right from the mobile device. The response of our text recognition API can be used to show the user or store in a database. In other cases there is an existing document flow, where orders are streaming in via mail or from an FTP location.
Whatever your current flow is, we can integrate. We work with customers with all sorts of needs. Sometimes we have to collect documents from a server or from a database. In other cases, we create a unique e-mail address where documents can be emailed to. To optimally benefit from the power of our OCR API, we recommend to use our web service directly via REST. Within a few seconds we will give you the result you are looking for.
OCR Output format and output stream
What has been just describe for input, also counts for the output stream. Our REST API for document recognition usually provides the response as JSON. Each web service can convert this output into fields in an external database. However, we can support any other format, when the custom flow requires that.
We have loads of experience with integrations of existing systems. Is your package only able to accept a .csv or Excel file as input? No problem, we store all recognized data as lines in a spreadsheet and provide you the output.
Because the original image often remains very relevant, we usually advise to use JSON or XML files in the document flow.
REST API for text recognition
Our OCR API come with clear documentation, which it is easy to read and then implement. Each RESTful API offers four options: GET, PUT, POST and DELETE.
The core of our API is the POST method for processing a document. The body of this POST is the document that you want to process, in a file format of your choice. It’s possible to add some additional information, for example a list of current suppliers, or order numbers, for us to help with matchmaking. The response contains a data object with all recognized data.
By using GET, general information about our API and the number of available credits can be asked for.
Would you like to have a conversation with us about the possibilities of our OCR API? We would love to get in touch and think along for the most ideal integration. Schedule a demo below with a product specialist or simply call us: +31 50 2111 631.