A comparison of multi-scale local binary pattern variants for bark image retrieval

Abstract

With the growing interest in identifying plant species and the availability of digital collections, many automated methods based on bark images have been proposed. Bark identification is often formulated as a texture analysis problem. Among numerous approaches, Local Binary Pattern (LBP) based texture description has achieved good performances. Bark structure appearance is subject to resolution variations which can be due to a number of factors (environment, age, acquisition conditions, etc). Thus it remains a very challenging problem. In this paper, we implement and study the efficiency of different multi-scale LBP descriptors: Multi-resolution LBP (MResLBP), Multi-Block LBP (MBLBP), LBPFiltering (LBPF), Multi-Scale LBP (MSLBP), and Pyramid based LBP (PLBP). These descriptors are compared on two bark datasets: AFF and Trunk12. The descriptors are evaluated under increasing levels of scale space. The performances are assessed using the Mean Average Precision and Recall\Precision curves. The results show that multi-scale LBP descriptors outperform the basic LBP and MResLBP. In our tests, we observe that the best results of LBPF and PLBP are obtained under low scale space levels. We also observe similar results for MSLBP and MBLBP across the six scales considered.