Nowadays, circuit boards are widely used in various fields, and solder balls on circuit boards may develop defects during the soldering process. Therefore, we need to find a way to inspect whether there are defects in the solder balls on the circuit board. There are various methods available for this purpose, such as ART (Algebraic reconstruction technique), SART (Simultaneous algebraic reconstruction technique), and FBP (Filter Back Projection). ART and SART are iterative methods that are more accurate but require more time. In contrast, FBP (Filter Back Projection) uses back projection to reconstruct solder balls and can save much time. In this experiment, we aim to obtain 269 slices of solder balls on the circuit board using FBP. We will stack all the images using ImageJ to generate a 3D image and probe whether there are defects in the solder balls. Afterward, we will employ different sorts of filters to conduct tests and use ImageJ for analyzing and comparing the characteristics of each filter. High-pass filters like the Ramp filter mostly enhance the edges of the image, whereas low-pass filters like the Hann filter mainly enhance the low-frequency components of the image, increasing contrast of the image. Following the analysis and comparison, we attempt to create our own filter.