All pipelines are subjected to corrosion and metal loss defects. Characterization of the defects is necessary to prevent environmental disaster due to leakage of oil or gas. Intelligent Pipeline Inspection Gauge (PIG) using MFL technique enables reliable inline inspection of pipelines. A high speed DSP based data acquisition system is designed for acquiring MFL data. An instrumented PIG acquires large amount of data from a trial run. Effective and real time compression requires fast processing of data using DSP. Only useful MFL data needs to be stored for analysis on a mass storage device. Multiresolution analysis of MFL data using discrete wavelet transform has been found most effective technique for this purpose Compression in wavelet domain gives more compression compared to other techniques. Although the technique is lossy, using wavelets can be shown that newer techniques of compression retain complete information for characterization of defects. A very high compression ratio of order of 1:20 or higher is achievable using this technique.