Image

Table of Contents

• Image matrix
• Image region

An image is a basic data structure, that represents visual or spatial information in a structured, pixel-based format. This document describes the image type used in the IDS peak ICV library and its associated components. It focuses on how images are represented, accessed, and processed. The goal is to provide a clear understanding of the image model to support correct and efficient usage. Both simple and complex image types are supported, including grayscale images, color images, and depth maps. The following sections explain the structure of the image matrix, the coordinate system, and the concept of image region.

In the IDS peak ICV library, images are represented using two key components:

Image Matrix: This contains the pixel data of the entire image. It's a two-dimensional array where each element corresponds to a pixel.

Image Region: This defines a specific area of interest within the image matrix. It allows operations to be performed on a subset of the image without modifying or referencing the entire matrix.

Image matrix

The image matrix is a two-dimensional array that represents an image, where each element corresponds to a single pixel. Each pixel can store:

• A scalar value (e.g., intensity, depth, confidence), or
• A multi-component value, such as color (e.g., (r, g, b)) or a 3D coordinate vector (e.g., (x, y, z))

This allows the image matrix to represent both visual and spatial information, making it suitable for applications like image processing or 3D reconstruction.

The image uses a two-dimensional coordinate system with the origin located at the top-left corner. Pixel positions are specified using (x, y) coordinates, where:

• x increases to the right
• y increases downward

The image has the following dimensions:

• Width: Number of pixels along the x-axis
• Height: Number of pixels along the y-axis

Image region

An image region defines an area of interest within the image matrix. Unlike a simple bounding box, the region is not restricted to a rectangular shape — it can represent any arbitrary subset of pixels, including but not limited to:

• A rectangular region
• A polygonal area
• A set of sparse pixel coordinates
• Or any shape defined by application-specific criteria

By default, the region encompasses the full image area. This ensures that standard operations apply to the entire image unless a more specific region is defined.

These characteristics apply:

• The region identifies the target pixels for processing or analysis.
• Pixel locations within the region are still addressed by their global (x, y) coordinates.
• The region must lie within the bounds of the image matrix.
• While operations produce results within the region, they may depend on pixels outside of it — for example, when applying filters or other neighborhood-based computations.

This abstraction enables flexible, efficient and focused processing of selected image areas — whether for segmentation, feature extraction, or targeted analysis.

Image Data	Full Region (Default)	Reduced Region

For instance, when applying a threshold, only relevant pixels are used, while pixels outside the region (considered irrelevant) are ignored.

Original image	Threshold on full region	Threshold on reduced region

Note that for most operations, the region of the input image is transferred to the output image. However, there are exceptions – some operations ignore the region or set it to the full image area in the output. Such operations are clearly marked in the documentation.

Image regions are stored in file when saving an image to disk, as long as the image file format supports it. For further information see the SaveToFile operation description.

For more information about a region see Region.