Digital-image dimension reduction via analysis of principal component

الملخص EN

A high-resolution image takes up more storage space, in addition to the data transit difficulty.

The Analysis of Principal Component, or PCA for a brief notation, is a mathematical approach utilized to lessen the data dimensionality.

It extracts the main pattern of a linear system using the factoring matrices technique.

The objectives of this paper are to see how effective PCA is in reducing digital picture features and to investigate the (feature-reduced) images' quality on comparison with different values of the variance.

As per the synthesizing of the initial research, the dimension or size reduction technique through the Analysis of Principal Component typically involves of 4-important steps : (1) picture-data normalizing (2) matrix of the covariance calculating using picture-data.

(3) discovering the picture-data projection (with fewer number of features) to a new basis use the Single Value Decomposition technique (SVD) (4) determining the picture-data projection (with fewer number of characteristics) to a new basis.

According to testing (500 picture) in language python and program anaconda3 results, the PCA approach considerably decreases the size of picture data while sustaining the original picture's fundamental properties.

This approach reduced file size by 35.3 percent for the best feature lowered quality.

The upload time of picture files through the Internet has substantially improved, particularly for mobile device downloads.