How to integrate recognition technologies into own software
The effective implementation of a solution of the data capture is sufficiently complex, much more than what you would normally expect.
Anyone who needs to use such technology usually choose products that are already tested and ready to use that minimize effort and maximize the benefits.
The best products data capture are in fact customizable enough to meet most needs.
However, for special needs, it may be necessary to carry out their own application alone: for this purpose it is possible use to the use of specific SDK (Software Development Kit ).
Conceptually, a SDK can be seen as a "black box" capable of images processing according to the instructions given, however, without the need to learn and develop the complex algorithms and technologies that are the basis of its functionality.
So a SDK consists of a library of functions available for the Windows platform as a DLL, of a documentation and some examples in different programming languages, such as Visual Basic, Visual C + +, Delphi, Visual C#.NET, Visual Basic. NET, Java.
The SDK therefore facilitate and speed up not just the realization of your project without the need to start from zero and collide with issues already resolved very well.
There are development Kit that implemented functionality of OMR, ICR, OCR, BCR, Deskew, Despeckle, Black Border Removal, Imaging, Identification Form, Form alligment, Dynamic Thresholding, etc.
View the availability of the SDK, for a user with an IT infrastructure capable of developing software, it is therefore advisable to buy a finished product ready to use or developed one at home?
It is not always easy to have clear ideas about the best way forward! Both possibilities have some pros and cons, but some reflection can help in making the right choice.
In some ways ita little like making the decision whether to have a bespoke suit rather than buying one already made.
A custom dress can dress to perfection better than one already prepared, but it is true that before it has been made you can not try wearing.
Similarly, a finished product can be viewed and tested before being purchased in order to verify whether and how to satisfy your needs, while a product to be made to measure, even though on paper reflects the requirements at the start, only after it realized it could be sure not to make a error or design implementative.
To get a bespoke suit is necessary to obtain the fabric, buttons, sewing machine, the accessories to make sure that everything you need is of suitable quality and in sufficient quantity is therefore necessary to have some familiarity with this activity.
Similarly to make an application for data capture must obtain all the SDK you need, without neglecting anything:
for example, even if it must read only boxes marking not only serves an SDK that is able to discriminate between boxes full or empty,
but also serves an SDK that is able to manage the acquisition of the images, the correction of the skew, cleaning, alignment with the reference template and do all those other essential tasks that are taken for granted and maybe you do not even notice in a readable solution ready for use.
The cost of a dress ready is known a priori, while the cost of a bespoke suit is only partially a priori estimated by adding up the cost of materials, equipment and labor: in fact, it is conceivable on past experience, but it is susceptible to variations according to the capacity of the "tailor" and contingent factors related to it
Similarly a data capture soluzione ready to use have a cost known at the start,
while for a measure to be implemented on addition to the cost of the SDK and other development tools, you will have to evaluate
also the cost of human resources needed to realize it, that, for obvious reasons
could also be understood in a very broad range of time.
A company that was the mission of make a clothes
on a large scale may invest in research and can take advantage of skills and differend experience to
arrive at finished product of high quality and free of defects, which however can not make a
craftsman who makes a dress of measure even though it has the raw materials and equipment valid.
Similarly, a solution of data capture
packaged is presumably the result of
substantial investments that those who must carry out a
own solution at home can not afford.
A dress you can wear it ready immediately,
while one made to measure can be wear
only after a long waiting:
must take the steps, you should baste, you have to sew, you must re-measure, you have to adjust, etc.
Similarly a data capture solution ready to use can go into production the day after buy,
while for a solution made at home you have to wait to be prototyped, coded, developed, tested, corrected, etc.
The balance seems to hang so much for the use of standard solutions rather than ready to use solution for home, while making use of the SDK.
But then, who can be interested in making their own solution to itself using the SDK?
In the first instance those who have very special needs,
for which you can not find a solution by data capture standard ready for use:
just as if you are very overweight, or have longer arms than normal,
or if you are extremely high or low and you can not find a dress that standard dress appropriately, and use the dress to measure.
In the second instance those who want to make an application of data capture extremely verticalised not to be used internally and distribute
to third parties in large numbers, unless of course, to find a valid commercial agreement with a manufacturer and therefore act as a dealer.
We can therefore conclude that usually more advantageous to use a standars solution of
data capture ready to use and customizable, rather than developed a solution at home, while
to the use of the SDK:
although in fact the SDK are indispensable and of great help,
however there can be no certainty of cost, time and above all,
result in creating complex application and complete as those related to the automatic extraction of data from forms and documents.
