Hello everybody, I am doing research in digital image processing. I want to know a few research questions on image processing.
Please give me some research questions or problem on image segmentation, enhancement or other latest challenges on digital image processing, which can be solved by mathematical analysis or Matlab programming.
You should consider the texture recognition. It is useful in detection of type of brain tumor with the MRI images. So without going into the surgical tests, you can diagnose the type of brain tumor.
Another direction is the image compression. You can work on the medical image compression. Because, now a days x-rays are to be stored on the memories which can take a lot of space. So you can apply loss-less compression to x-ray image without compromising the specific part of it where there is the information of the decease and remaining part of x-ray can be compressed using lossy compression methods as those areas are insignificant from the diagnostic point of view.
You should consider the texture recognition. It is useful in detection of type of brain tumor with the MRI images. So without going into the surgical tests, you can diagnose the type of brain tumor.
Another direction is the image compression. You can work on the medical image compression. Because, now a days x-rays are to be stored on the memories which can take a lot of space. So you can apply loss-less compression to x-ray image without compromising the specific part of it where there is the information of the decease and remaining part of x-ray can be compressed using lossy compression methods as those areas are insignificant from the diagnostic point of view.
Hi, Kuldip, here you are some interesting challenges for digital image processing:
texture analysis and description (very useful for image segmentation), you could study in deep this argument with this paper that investigates the scale of texture. This is the link to the paper http://www.researchgate.net/publication/256801807_Scale_detection_via_keypoint_density_maps_in_regular_or_near-regular_textures/file/72e7e523c678cd01e2.pdf
Another interesting challenge for digital image processing is the visual saliency that deals to fix the most important points of an image from a perceptive viepoint, this could be useful for an advanced inspection of the digital image that is the analysis of the most important regions of an image. (Here, i post the link to my visual saliency method: https://www.researchgate.net/profile/Giuseppe_Mazzola/publication/236233161_Visual_Saliency_by_Keypoints_Distribution_Analysis/file/e0b495173c30c9bf18.pdf?origin=publication_detail).
Another aspect that is very important and very current in the state of the art of the digital image processing is the analysis of the keypoints of the image such as SIFT, SURF, ORB, ASIFT. This points (accompained by descritpors vectors) are very important to represent an image and in the filed of image and object matching. Here a link to an Image Forensic method that makes use of SIFT keypoints descriptors: https://www.researchgate.net/publication/221127756_Detecting_multiple_copies_in_tampered_images .
I hope that this suggestions could be helpful for your future works:
Good Luck!
A.B.
Article Scale detection via keypoint density maps in regular or near...
Conference Paper Visual Saliency by Keypoints Distribution Analysis
Conference Paper Detecting multiple copies in tampered images
An OCR (Optical Character Recognition) system would be a good area to work on. You may work on PDE based image segmentation techniques.
You can work on any field of Image Processing according to your interest but you can use Hexagonal Image Processing in your work for better results, that's why explore Hexagonal Image Processing please
Hi Kuldip, first you be specific about the area of your interest. Image processing can be implemented to any field, be it a medical image or satellite image or texture etc. The image processing techniques required may differ with respect to the type of image.
Remote sensing is very interesting area in image processing.. You may work object-based image segmentation techniques on remote sensing.
Content Based image and video Retrieval is also an challenging and more upcoming research topics
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing
- Image Processing