Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 8 years ago.
Improve this question
Hough Transform Algorithm is one of the algorithm use for text line detection and edge detection.
Does Hough Transform Algorithm can be use for Detecting text in Images?
What must be the process or implementation in java for this question? or It must be have another algorithm to make it possilble?
Hoping for positive response.
Hough transform was initially designed to detect straight lines in an image, but later modifications have been proposed to use similar logic to detect circles or other shapes. However letters and text is a bit more complex and can(most of the time) be described with a simple shape. Thus applying Hough transform directly is not possible.
However Hough transform can be used as part of the solution. For instance you can detect straight lines and this can help you detect separate letters or to recognize some special letters. Also sometimes Hough can be used to detect text direction which is a challenging problem on its own.
Text recognition is a very interesting problem and many different solutions have been proposed. In fact there are whole courses dedicated to that and new approaches appear often. I personally like approaches that use neural networks the most. Have a look here for instance.
Related
Closed. This question needs details or clarity. It is not currently accepting answers.
Want to improve this question? Add details and clarify the problem by editing this post.
Closed 1 year ago.
Improve this question
I'm fairly new to OpenCV Java and I've been looking for methods to do what the title asks.
So far, from what I understood, the first step should be to use a detecting edges filter like Sobel or Laplace.
Coming from https://docs.opencv.org/3.4/d2/d2c/tutorial_sobel_derivatives.html, I get the following result : glass_sobel
From now on, I don't know what to do next... What should I do to get what I want ?
Thanks for your help !
Taking the vertical edge difference (sobel) is indeed the right step. But first I suggest fixing some parameters. For instance, is the position, size and shape of the glass fixed? then take an image without fluid and use that to extrapolate the glass from the result. If not, I would take the following steps:
Try to increase edge visibility. In your image you do see a clear edge, but the computer might see this as 4 edges, therefore try to make sure edges close to each other are seen as one edge (dilate operation might do that). Depending on the liquid, you could alternatively extract a certain color from the image. Water is here the most unfortunate case, but something as orange juice is easily detected.
With the more clear edges, I would try to get a vertical line out of the glass, and classify each edge (is it part of glass or part of liquid).
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 6 years ago.
Improve this question
Contex: I'm trying to make a 2d physic engine, i'm starting to think about how to draw shapes that i could rotate. So i thought i could draw irregular shapes by drawing lines from several equations of straight. Now i have some doubts:
There is a easy way of draw a straight by giving an equation?
There is a library that can help me to handle equations?
The reason to use equations instead of using functions to draw lines from x1,y1 to x2,y2 is that i want the equations to calculate collisions between shapes.
Do you know a book or an article that could help me pull this off?
I'm gonna work in java, for now this is just an idea.
Not an expert in this field but quick search on java 2d engine collision engine returns dyn4j library on the first page. You can explore their examples and consequently the implementation code.
https://github.com/wnbittle/dyn4j
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 6 years ago.
Improve this question
I am new to the image processing discipline, currently I am working on an activity to find the path in a given floor plan image using java. I could understand & implement a plain Dijkstra or A star algorithm in java to find the shortest path between nodes that can work on hard coded values for nodes & edges. But I have no idea on how to do the same with image file as input.
I could not find any much detailed or comprehensive solution for this on web surfing. Any idea on how to achieve this. Kindly help.
If the walls on your plan are black, and the space is white, then use the white pixels as the space in which you can search using your algorithms.
be aware that:
The plan images could need some preprocessing, like adjusting brightness/contrast and/or converting into lower resolution picture.
If the plans are more complicated than that, ie. you need to take into account some special signs or structures on them then you need to do more complicated conversion/preprocessing and/or image recognition, using some special tools, maybe even some AI.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 7 years ago.
Improve this question
I'm having trouble understanding the concept of using FFT to calculate frequency changes. I've been scavenging through a ton of articles on this concept but it's still confusing me. I found a Stack Overflow article about the steps to take to calculate the frequency, but I'm still a little bit confused on exactly what to do in Java.
Sounds like you need to start way back with some general concepts. An FFT is just a fast DFT. A DFT uses Fourier decomposition on a set of sampled data. So first you need to understand what a waveform is, what a sinusoid is, what a frequency is (and is not the same as a musical pitch), and how Fourier's theorem works to relate the two. Then how to compute a DFT on a window of a sampled waveform, and how to interpret the complex result vector. Then how to do it fast in Java.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 8 years ago.
Improve this question
How to look for a particular object in a large list(java collection) without iterating it. Assume we have a large collection and just need to check whether a object exists or not without iterating.
Ok, Let's step out of the binary world.
Think of a chest full of Lego parts. You want a 2x2 flat black piece.
How would you find it without looking in the chest?
There is no magical to find it, you need to jump into the chest and find the piece grabbing one by one and checking if it's the one you are looking for.
There are ways to speed up the process.
You can Organise (sort your collection) by colour for example and just look in the black pile.
Or you can map (Index your pieces) so you know the position of the piece and can go and retrieve from you know where the piece is.
That is, in a very simplistic way, the same idea for databases and collection.
So, summarizing, no, you can't not just find without looking. Sorry :(