Abstract：
When developing detection techniques for fingermarks, the detected fingermarks must be evaluated for their quality to assess the effectiveness of the new method. It is a common practice to compare the performance of the new (optimized) technique with the traditional or well-established ones. In current practice, this evaluation step is carried out by a group of human assessors. A new approach is applied in this paper and consists of using algorithms to perform this task. To implement this approach, the comparison between IND/Zn and DFO has been chosen because it has already been the subject of many articles published in recent years and a consensus exists on the superiority of IND/Zn over DFO. The quality of 3\'600 fingermarks developed using both detection techniques was assessed automatically using two algorithms: LQM (Latent Quality Metric) and ILFQM (Improved Latent Fingerprint Quality Metric). The distribution of quality scores was studied for both detection techniques. The results showed that fingermarks detected with IND/Zn received higher scores on average than fingermarks detected with DFO, which is in line with the consensus in the literature based on human assessment. The results of this research are promising and shows that automated fingermark quality assessment is an efficient and viable way to comparatively assess fingermark detection techniques.
摘要：
在开发指纹检测技术时，必须评估检测到的指纹的质量，以评估新方法的有效性。通常的做法是将新的(优化的)技术的性能与传统的或成熟的技术进行比较。在目前的实践中,该评估步骤由一组人类评估员进行。本文应用了一种新方法，该方法包括使用算法来执行此任务。为了实施这种方法，选择IND/Zn和DFO之间的比较是因为它已经成为近年来发表的许多文章的主题，并且对于IND/Zn优于DFO存在共识。使用两种检测技术开发的3'600个指纹的质量使用两种算法自动评估:LQM(潜在质量度量)和ILFQM(改进潜在指纹质量度量)。研究了两种检测技术的质量分数分布。结果表明，用IND/Zn检测到的指纹平均得分高于用DFO检测到的指纹，这符合基于人类评估的文献中的共识。这项研究的结果是有希望的，并表明自动指纹质量评估是比较评估指纹检测技术的有效和可行的方法。
