Approach to Object-Based Embedded Coding with Optimized Truncation

Traditional EBCOT algorithm encodes an image with the same block-based coding strategy and it is difficult to acquire perceptually consistent results. This paper presents an improved object-based embedded coding algorithm to enhance the flexibility and coding efficiency of the original EBCOT algorithm. The original image is segmented into objects and each object can be encoded independently with different strategy according to its visual interest. At the decoder, each object can be reconstructed progressively depend on the given bit-rate. As a whole, the improved algorithm is more flexible and may enhance the overall subjective image quality. Experiments are conducted and the results show that the proposed algorithm is effective.