Skip to main content

Image Recognizer App (WIP)



This was done for image identification of any particular classification of images that are identified by a script trained on data sets and has the intelligence to identify a good number of objects and give generic info about it.
The overview of the idea is to create an image that captures an image and delegates it to a server where the script(ML trained script) is present and get the processed result and publish it to the user. The delegation method is via a rest call from the app to the server port where the script is running.

The current plan is to run a python flask app in a certain port in the localhost and make Rest calls from the app(In the form of threads that run in the background.) and process information obtained from the server and store it in the database.

Step 1: Building a simple hello world flask app.

Step 2: RestFul API supported android application.
Watching a tutorial on lynda.com - https://www.lynda.com/Android-tutorials/Integrating-Android-Apps-RESTful-Web-Services/560057-2.html

I didn't have a subscription to lynda.com. But there is a workaround to get access to a good set of tutorials for free.
You can go ahead and pay the subscription and use the services. But just in case you want to try the workaround. Please watch the following video and create a free subscription account on lynda.com
Link: https://www.youtube.com/watch?v=P9iPbB0vmNA

Note: Started working on it and currently put on hold due to other commitments. Will write follow up update posts when I get back to working on it. Until then, see you. Happy coding! :) 

Comments

Popular posts from this blog

The one where Raj went for Cycling (First time Experience - 50km circuit)

The Past: I met a friend Srini in violin class and got to know that he goes for cycling with a bunch of explorers once in a month. I was excited and told him I had a cycle but haven't done cycling and would love to try. So he told me he will inform when they go for a ride. The Present:  Srini texts me that they were planning a morning ride of approx 50-60km circuit. Even though I was excited, I was a bit nervous if I could push myself for that far. I did think of it for a while and I was like "If not now, then when??? ". I said yes. It was a Sunday morning that we started off. The weather in Bangalore was a big threat as it was raining pretty consistently for the past fortnight. We trusted few weather reports which stated that there is not going to be any rain in the morning and there might be slight rain in the afternoon. Srini started from HSR, two of friends joined in Bellandur, I joined in Kadubeesanahalli junction and two more joined in ISRO junction. The r

Implementing a file monitoring script using python.

The idea is to implement a method where a user will be able to modify an xml file and the modifications will trigger a certain action in the script. This has to be implemented in python. WatchDogs is a python library that helps in file system monitoring. API reference : https://pythonhosted.org/watchdog/quickstart.html#quickstart Useful link :  http://brunorocha.org/python/watching-a-directory-for-file-changes-with-python.html The example scripts provided in the above links give a good headstart to play around with the file system and figure what is the exact use case that you want to achieve. Current use case:  The python script that I run is bundled as a container. There are mounted volumes in the container. So using an external script I am planning to modify the xml file and since the same file is modified inside the container, the python thread that is listening to the file modification event will trigger an action like sending a mail, make an entry to the log file.

Good programming practices

Over the past few days, I have been pondering over the fact that despite having a good hold on OOPs concepts and apply them while writing code to solve a particular use case, I am still not writing good code. I have been coding for almost 6 years now. 2+ years of professional coding. When I was in college, I was a wild programmer solving programming contests, DS problems. What I mean is that I only coded to solve the problem, did a lot of dirty coding. I never bothered if the code can be reused, if someone else or I myself can understand it if I looked at it after a few days. I did have a good habit of writing comments.  When I joined work, I learned many different aspects of how the code was written. Many facts were taken into consideration like code maintenance, reusability, testing etc. I still remember a co-worker who told me if you need to write 10 lines of beautiful code, you should have seen 1000s of lines of amazing code. Due to the pressure of getting things done,