List of taxonomy terms

'Active contours' is an energy-minimizing deformable parametric contour, that is typically used to segment image regions enclosed by edges.

Parametric (global) registration restricted to affine transforms (allowing translation, rotation, scaling, shear, reflection and their compositions).

Re-occurs among biii.info tags (visualisation, rendering, viewer, classification, ...)|ToDo: Clear up relation between Classification, Clustering, and Prediction and recognition (also in main EDAM)

Information, represented in an information artefact (data record) that is 'understandable' by dedicated computational tools that can use the data as input or produce it as output.

Additive

multiplicative

Dilation is a morphogical operation in which an object's size increases by the radius of a structuring element.

Relation with Feature enhancement

Relation with Edge detection...

Filament tracing operations are image analysis operations in which there is an image of a filamentous structure (it may be a tree-like structure, a filament network or a agglomeration of single 'stick-like' filaments) as input and outputs data that represent the filament, most commonly a skeleton representation of the filaments and their diameters or surfaces.

https://en.wikipedia.org/wiki/Fourier_analysis https://en.wikipedia.org/wiki/Wavelet|Frequency domain analysis (or spectral density estimation), is the process of decomposing a complex signal into simpler parts. Many physical processes are best described as a sum of many individual frequency components. Any process that quantifies the various amounts (e.g. amplitudes, powers, intensities, or phases), versus frequency can be called spectrum analysis. The Fourier transform of a function produces its frequency spectrum by decomposing the function into its sine and cosine components. Whereas the standard Fourier transform is only localized in frequency, wavelets are localized in both time and frequency. Wavelets are handcrafted to correlate with particular frequencies in a signal; a set of wavelets is usually used for a complete analysis.

https://en.wikipedia.org/wiki/Fourier_analysis https://en.wikipedia.org/wiki/Wavelet|Frequency domain analysis (or spectral density estimation), is the process of decomposing a complex signal into simpler parts. Many physical processes are best described as a sum of many individual frequency components. Any process that quantifies the various amounts (e.g. amplitudes, powers, intensities, or phases), versus frequency can be called spectrum analysis. The Fourier transform of a function produces its frequency spectrum by decomposing the function into its sine and cosine components. Whereas the standard Fourier transform is only localized in frequency, wavelets are localized in both time and frequency. Wavelets are handcrafted to correlate with particular frequencies in a signal; a set of wavelets is usually used for a complete analysis.

Suggested def: image annotation is the process of defining metadata to a digital image, e.g. defining regions, marking points, creating textual descriptions, attaching tags to image contents. Can be manual or automatic.

Image correction is the removal of aberrations originating from the imaging instrument, lighting, or sample.

So far consensus: Primary term Image data, exact synonym Image, meaning image or a set of (related) images (forming e.g. a stack or a time-lapse). No synonyms for stacks and movies, these should be specialised concepts.

Image enhancement is a term used to refer to an operation that increases the quality of the image, being historically more used in the context of contrast enhancement (which improves visualization for the human eye). However, image enhancement is also a very general term that refers to operations that enhance features interest in the image. An image enhancement operation moves further away from the reality, as opposed to Image reconstruction that moves closer to the reality.

Image enhancement moves further away from the reality, as opposed to Image reconstruction that moves closer to the reality.

Image enhancement moves further away from the reality, as opposed to Image reconstruction that moves closer to the reality.

Image registration is the process of transforming different sets of data into one coordinate system. Registration is necessary in order to be able to compare or integrate the data obtained from different sensors/imaging modalities, at different times, from different view points, etc. . Registration can be based on correspondence established between the landmarks or feature points. Alternatively, some similarity/distance metric is established between the image intensity maps to navigate the registration process.

Image registration is the process of transforming different sets of data into one coordinate system. Registration is necessary in order to be able to compare or integrate the data obtained from different sensors/imaging modalities, at different times, from different view points, etc. . Registration can be based on correspondence established between the landmarks or feature points. Alternatively, some similarity/distance metric is established between the image intensity maps to navigate the registration process.

Image segmentation is (one of) the (few) concept(s) on the border between Image (pre)processing (Image->Image) and Image analysis (Image->Data).

Image segmentation is (one of) the (few) concept(s) on the border between Image (pre)processing (Image->Image) and Image analysis (Image->Data).

Visualisation vs Plotting vs Image generation. Should these be merged? Which of these should be the top concept, and which sub-concepts, and which narrow synonyms?

Labelling? Not allowed to stand alone, needs additional modifiers as means various things. Maybe not that relevant here. Anywhere else?

Concrete algorithm?

ToDo: add a few more but not too many specialised concepts under here.

Utilizing different imaging modalities in combination

A function that processes a set of inputs and results in a set of outputs, or associates arguments (inputs) with values (outputs).|Special cases are: a) An operation that consumes no input (has no input arguments). Such operation is either a constant function, or an operation depending only on the underlying state. b) An operation that may modify the underlying state but has no output. c) The singular-case operation with no input or output, that still may modify the underlying state.

Visualisation vs Plotting vs Image generation. Should these be merged? Which of these should be the top concept, and which sub-concepts, and which narrow synonyms?

ToDo: Clear up relation between Classification, Clustering, and Prediction and recognition (also in main EDAM)

Needs a comment about the relation between the Watershed and Region growing.

Resolving the discussion on object vs region vs region of interest, after adding Wiki def. Object is a region, but region can be more than an object.

Parametric (global) registration restricted to rigid transformations (translation and rotation).

Relation to Opening and Closing, and differences between top=white and bottom=black The top-hat transform: the opened image is subtracted from the original, and desired detail is obtained. The bottom-hat transform: the residual of a closing compared to the original image.

A category denoting a rather broad domain or field of interest, of study, application, work, data, or technology. Topics have no clearly defined borders between each other.

Re-occurs among biii.info tags (visualisation, rendering, viewer, classification, ...)

Watershed is the term that commonly refers to a mathematical morphology operation that treats a grayscale image as a topographic map and segments the image. The segmentation is performed by a succesive 'flooding' operation from minima in the image starting from different points and separates the image in different catchment basins.|Needs a comment about the relation between the Watershed and Region growing.