How can i determine the location of LCD/LED display in an image with OpenCV and perl

edit

Thank you for providing some example images. If they look similar to those you can identify your lcd-region by color since it is the only one which is in a yellow area.

Steps I would do:

1. clean ore denoise the source-image (e.g. by using cv::fastNLMeans()) or blur your image using a Gaussian-kernel: cv::GaussianBlur()

2. optional: normalize contrast --> cv::equalizeHist()

3. Iterate through your images and identify yellow pixels (a simple if-condition here)

4. get the bounding box of all these yellow pixels, i.e. get the min, max x and y-locations of your yellow points (you could do this also in the for-loop of step 3.)

5. Pass the region of interest of step 4 to the open-source ocr software tesseract (http://code.google.com/p/tesseract-ocr/) to obtain an ocr result.

If your LCD is not always looking yellow-ish then you have several options, here are some ideas:

1. identify the lcd by context, maybe you know that the lcd-line is always the second one in your image then you could also give tesseract the whole image and get an ocr of the whole image and the second line is then your result

2. if you know it is the largest bounding box, apply the first two steps from above and find connected components (cv::findContours()) then pass the largest bounding box of connected components to tesseract.

3. more advanced: if you know that it has the fattest stroke lines you could get the thickness by computing the distances between gradients of the strokes

4. also advanced: if this was really just an example image and all your others look completely different then you need a learning approach, e.g. you could try out the cascade-classifier of OpenCV

Well unfortunately I haven't worked yet with low level computer vision using OpenCV (I made that with MATLAB) so I can´t give you specific advice on functions, I have here a link talking about the morphological operations: LINK

But let's explain a little more:

In order to use the morphological (or blob) operation we need to convert the image you have (in color) to black and white (not gray scale) in order to do this you should choose a threshold (it could be more than one depending of the needs).

The point is that after analyzing the histogram you could search for a "mode" corresponding to the yellow rectangle (or even the numbers if you take only the higher part of the image). Check this: LINK

Then you could make use of the blob operations.

Hope it helps.

Supposing you use always the same kind of card, you could probably by bounding/restraining the problem and work to do.

For example you can make that the card position and size (for the image) be more or less constant, so you could take for example just the upper part of the image.

Then you should segment the display.

This can be made by using some morphological operations (dilation or closing) to get a mask and the apply this mask to cut the important part for this. If you scratch all the card (so there aren't any near black points close to the number) this could work very well.

Another one could be segmentation using the histogram of the picture.

Hope it could help.