5.15.2. Multispectral Images

A multispectral image may consist of several bands of data (3-10 bands), such as from Landsat TM, where each band image has a specific utility. Human eyes can’t appreciate the grey level variation of objects in each image, but there is variation which can be detected by the software. For visual colour display, each band of B&W image is displayed in one of the primary colours and superimposed. Thus, a colour composite image is much appreciated by human eyes for identifying many features/objects, as compared to individual B &W images.

Multispectral images are images captured in multiple wavelength bands, typically between 3 to 10 bands. Each of these bands captures different data about the surfaces on Earth, which can provide unique information about various features. Human vision can't fully discern the subtleties of these bands; hence, specialized software processes the data. When the images from these bands are combined, they produce a color composite image, making it easier for us to see and interpret the features within the data compared to black-and-white images.

Consider a gardener who wants to understand the health of plants in her garden. If she only uses a plain black-and-white photo, she might miss important details. However, using colored images, representing different information (e.g., healthy plants in green, unhealthy in brown), allows her to quickly identify problem areas and take action. This is akin to what multispectral images do using different bands of data.

The interpretation of a colour composite image however will require the knowledge of the spectral signature of the objects present in the scene.

Interpreting color composite images requires familiarity with the spectral signatures of various objects. A spectral signature is like a unique 'fingerprint' for different materials, capturing how they reflect light across different wavelengths. Knowing the spectral signatures helps analysts to distinguish between different land covers, vegetation types, or even soil conditions in the multispectral image.

Imagine a chef who knows how to identify ingredients by their colors and smells. If she wants to cook a specific dish, she can quickly determine which ingredients she should select based on their 'signature' qualities. Similarly, remote sensing experts use their knowledge of spectral signatures to identify and analyze different elements in multispectral images.

In displaying a colour composite image, basically three primary colours (red, green and blue) are used. When the three primary colours are mixed in various proportions, they can produce different colours in the visible spectrum.

Multispectral images use three primary colors (red, green, and blue) to create a composite image. Each color is assigned to specific wavelength bands corresponding to different features in the image, allowing for detailed visualization of the Earth's surface. By manipulating these colors, analysts can highlight differences in land cover and other features effectively, making interpretation more straightforward.

Think of a painter mixing different colors to create a vibrant landscape. By combining red, green, and blue paint in various amounts, they can depict trees, skies, and flowers accurately. In multispectral imaging, scientists mix data from different bands similarly to visualize the Earth's features in a way that is colorful and informative.

If the colour of an object in the colour composite image may not have any resemblance to its actual colour, the output image is known as a False Colour Composite (FCC).

False Color Composites (FCC) are produced when the colors displayed in an image do not represent the true colors of the objects. Instead of showing what we see with our eyes, FCC highlights different features based on their spectral properties. This technique can sometimes create confusion unless the viewer understands what the colors represent, which can limit the straightforward interpretation of images.

Imagine someone wearing a pair of colored glasses that alters what they see; reds appear blue and greens appear yellow. While they’re still looking at the same objects, their perception of the colors has changed, potentially leading to misunderstandings. Similarly, in FCC images, understanding the representation of different colors is crucial for correct interpretation.

--