cv2imread

cv2.imread, often referred to as cv2.imread in Python, is a function from the OpenCV library used to load an image from disk into a NumPy array. It is part of the OpenCV-Python interface and serves as the primary way to bring image data into a program for processing.

The function reads the file and returns a multi-dimensional NumPy array representing the image data. If the read fails (for example, due to a missing file or an unsupported format), it returns None. By default, cv2.imread loads images in color, using the BGR channel order rather than RGB.

Parameters:

- filename: the path to the image file.

- flags: optional; determines how the image is read. Common values include:

- cv2.IMREAD_COLOR: loads a color image (default). Any existing alpha channel is discarded.

- cv2.IMREAD_GRAYSCALE: loads the image in grayscale as a single-channel array.

- cv2.IMREAD_UNCHANGED: loads the image as-is, preserving any alpha channel.

- cv2.IMREAD_ANYCOLOR and cv2.IMREAD_ANYDEPTH are also available for broader loading options in some contexts.

Return value:

- A NumPy array with shape (height, width, channels) for color images (typically 3 channels in BGR order) or (height, width) for grayscale images.

- None if the image cannot be read.

Common usage tips:

- Always check the result for None before proceeding.

- If color is required in RGB order for display or processing with other libraries, convert with cv2.cvtColor(img, cv2.COLOR_BGR2RGB).

- Be mindful that OpenCV loads images in BGR by default; many image viewers assume RGB.

See also: cv2.cvtColor, cv2.IMREAD_GRAYSCALE, cv2.IMREAD_UNCHANGED, cv2.imdecode.