The edge detection takes a wide interest in the field of digital image processing and computer vision, which is one of the basic stages in the process of patterns recognition and image segmentation. In Digital image processing, a set of methods and algorithms are applied on data of digital images in order to obtain some basic information that can be used in a particular application such as image enhancement, image segmentation , edge detection and others. The image processing algorithms have been updated as a result of the development in computer technologies such as processing speed and ease of computer acquisition in addition to the development of cameras. Digital images consist a set of elements (pixels) formed as a two-dimensional matrix the edge is a narrow area produced between two different regions in the values of brightness or difference between object and background. Edges detection methods have most important in the processing of digital images, which is a key stage in the distinction of patterns. Most edge detection methods high pass filter pass the image and mathematical convolution between filter and image.
Papers:
Tiwaskar 2013 "Comparison Of Edge Detection Techniques", Proceedings Of Sixth International Conference 2015.
Mingfeng Zhu, Jianqiang Du, and ChenghuaDing, 2014: "A Comparative Study of Contemporary Color Tongue Image Extraction Methods Based on HSI", International Journal of Biomedical Imaging.
I agree with you that most of the edge detection efficiency models are based on gray level images. That is due to the nature of edge detection techniques. However, to overcome this limitation, other segmentation techniques can be used to reduce n color space to a dimension which can be easily handled by edge detection techniques and can be easily evaluated by efficiency models
Most of the edge-detection methods based on colors and layers are application specific. Application - specific means is what type of illumination one has when forming the image (i.e. depending on the image sensor, being broadband-light based, narrow-band based, photographic-like medical imaging; artificially-coded imaging as in PET imaging). In addition to that, you are talking about efficiency models - that means - this is also in relation to coding procedures and if the raw image can be transformed into specific type of imaging (JPEG for example) and if the compressions or coding uses some type of redundancies. Therefore, the effectiveness and the color-specificity are oprimisation parameters that can be researched independently or, in combination.
i m working in the field you may find the various methods but when it comes to human perception, true positive true negative ,false positive ,false negative are most reliable merits. They might be more time consuming but way more accurate.
- Badges
- Science topic
- Similar topics
- Computer Science
- Images