It is effective against noise and loss of information. It increases the accuracy of classification as well as decreases computational time.

Advantages:

1. Only object/class-specific information is used for classification. This reduces the impact of background-to-background matching.

2. May improve performance when the object is very very small.

3. Can sometimes reduce computation, if you're performing segmentation and classification as individual sequential tasks. Particularly, when the object is way too small and you crop after segmentation.

Disadvantages

1. Removes contextual information. (Say while classification, you cannot find monkey class in the ocean and a ship on a tree). Hence, the background information can sometimes help in giving a few relevant details. This point vs point #1 of advantages depends on the problem statement.

Segmentation increase classification accuracy, since it helps extract features in an image. However, it is different from feature selection which has a better impact on classification.

It is effective against noise and loss of information. It increases the accuracy of classification as well as decreases computational time.

Dear Md. Muhaiminul Islam

Check these articles. They will directly answer your question.

Article Optimized superpixel and AdaBoost classifier for human therm...

Article Breast Cancer Segmentation From Thermal Images Based on Chao...

Segmentation of images is the process which helps to segregate images in different classes for its indexing, classification, identification and retrieval for efficient image processing and management. Image segmentation is the process of partitioning an image into different segments as image objects for its efficient object classification, identification for image understanding, interpretation and analysis. Intensity variation of pixels plays a vital role in image segmentation. Segmentation can be performed at color, pixel, pit-plane levels.

Segmentation is in fact a type of classification. Ideally, each image segment (blob) should be labeled with a single class. However, there are other ways to apply a class without segmenting the image into blobs. These classifiers are typically implemented in black-box approaches using Artificial Neural Networks.

Also have a look at https://neptune.ai/blog/data-exploration-for-image-segmentation-and-object-detection

https://www.analyticsvidhya.com/blog/2019/04/introduction-image-segmentation-techniques-python/

Segmentation is usually done to isolate the object of interest from the background so that visual features like color or texture are computed from relevant portions only. Segmentation might also be used to identify the boundary or contour of the foreground object in case of shape based classification.

Its a process of assigning a label to the pixel of image. As same label pixels contain absolute feature.