File size bounded JPEG transcoder (FSBJT)

摘要: A technique that reduces the size of an existing JPEG file or set DCT coefficients to satisfy a certain bit budget by setting zero whose magnitude is below threshold and which occur after ordinal number in zig-zag scan. The cutoff chosen using clever savings calculation strategy. This strategy implemented filling appropriate values array values, Savings[1], . , Savings[63]. value Savings[n] exactly bits saved reducing thresholding from n +1 n. When non-zero coefficient zero, are because two runs zeros (the one preceding it following it) get combined into single, longer run zeros. exact can be calculated adding needed code previous next runs, subtracting run. Some special conditions (runs than 16 end-of-block conditions) need carefully factored computation.