Analysis and Improvement of JPEG Compression Performance with Discrete Cosine Transform and Convolution Gaussian Filtering

Although computers and smartphones now have graphical capabilities for conveying data information, there are limitations on the size of the file used to store the data. As a result, it takes longer to transmit data to another location, which in turn affects the required time. This paper presents a compression method that utilizes the Discrete Cosine Transform to minimize the number of coefficients needed to represent the input data. The quantization and masking processes then remove relatively small coefficients without causing visual distortion while maximizing compression and maintaining image quality. Furthermore, post-processing techniques are employed to improve the quality of compressed or degraded images resulting from the DCT transformation process. The Gaussian Filter method is used to smooth the image and remove normally distributed noise. Consequently more compact information is obtained, which reduces the transmission time when it is sent to another location.