| Author Name | Affiliation | | Babar Khan | College of Information Science and Technology, Donghua University, Shanghai 201620, China | | Fang Han | College of Information Science and Technology, Donghua University, Shanghai 201620, China | | Zhijie Wang | College of Information Science and Technology, Donghua University, Shanghai 201620, China | | Ather Iqbal | College of Information Science and Technology, Donghua University, Shanghai 201620, China |
|
| Abstract: |
| An automatic intelligent system for the colour and texture inspection of bakery products is proposed. In this system, advance classification technique featuring Support Vector Machine and biologically inspired HMAX based shape descriptor integrated with biologically plausible RGB Opponent-Colour-Channel Descriptor is used to classify bakery products to their respective classes based on the shape and based on their colour referring to different baking durations. The results of this paper are compared with other methods for the automatic bakery products inspection. It is discovered that biologically inspired computer vision models performs accurately and efficiently as compared to the computer vision models which are not biologically plausible, in the bakery products quality inspection. It is also discovered that the One Versus One SVM and Directed Acyclic Graph SVM acquired the maximum accurate classification rate. The proposed method acquired classification accuracy of 95% and 100% for the biscuit shape and biscuit colour recognition, respectively. The proposed method is also consistently stable and invariant. This shows that the biologically inspired computer vision models have the capability to replace existing inspection methods as more reliable and accurate alternative. |
| Key words: Bakery products quality inspection computer vision HMAX opponent color channel RGB color descriptor Support Vector Machine (SVM) |
| DOI：10.11916/j.issn.1005-9113.16179 |
| Clc Number:TP391.41; TP183 |
| Fund: |
|
| Descriptions in Chinese: |
| 基于形状和颜色信息的烘焙产品的自动质量检测 Babar Khan，韩芳*，王直杰，Ather Iqbal （东华大学 信息科学与技术学院，上海 201620） 创新点说明： 利用支持向量机和基于生物启发式HMAX模型的形状辨识器与RGB对抗颜色通道辨识器集成而形成一种有效的智能分类系统，并将其用于烘焙产品的质量检测。 研究目的： 本文要对烘焙产品进行更有效的质量检测，需要提出一种智能方法。 研究方法： 本文采用了支持向量机、基于生物启发式HMAX模型的形状辨识器与RGB对抗颜色通道辨识器集成，形成一种高级自动分类系统，对烘焙产品根据其形状和颜色（烘焙时间）进行对应分类。 结果： 利用本文方法对烘焙产品进行质量检测的结果与其它方法相比较，发现本文提出的基于生物启发的机器视觉系统运行更加精确和有效，同时发现采用One Versus One支持向量机和Directed Acyclic Graph支持向量机可以得到最大的分类精确率。本文方法对饼干形状和颜色的辨识分别达到了95%和100%的分类精确率。同时，算法稳定。 结论： 本文方法因其更加可靠和精确可用于代替现有的烘焙产品检测方法。 关键词：烘焙产品质量检测，计算机视觉，HMAX, 对抗颜色通道，RGB颜色辨识器，支持向量机 |
