摘 要：针对某些应用中目标表面纹理较少，目标检测困难的问题，提出了一种基于区域梯度压缩的少纹理目标候选框提取算法。该算法是对模板匹配算法OCM 的改进。算法对局部区域梯度方向进行压缩，保持了较低的计算复杂度，并且提出了新的梯度方向压缩方法与相似度衡量方法。实验证明，该算法相较于OCM 算法，在产生接近数量候选框的情况下，召回率提高了6.5%；在召回率接近时，产生的候选框数量减少了41.9%。
关键词：少纹理目标；目标检测；模板匹配；目标候选框提取；量化编码梯度方向；二进制梯度方向压缩
中图分类号：TP391 文献标识码：A 文章编号：2096-4706（2020）02-0102-04
Texture-less Object Candidate Box Extraction Algorithm Based on Region Orientation Compression
PENG Maoting
（College of Electronic and Information Engineering，Nanjing University of Aeronautics and Astronautics，Nanjing 210016，China）
Abstract：In some applications，the objects to be detected don’t have enough surface texture informations，which causes a great challenge to the accurate object detection. Aiming at above issue，this paper proposes a candidate box extraction algorithm for texture-less object based on region orientation compression. This algorithm is an improvement on the template matching algorithm OCM. The algorithm compresses the local area edge orientation and maintains a low computational complexity，the algorithm proposes new orientation compressing method and similarity measurement method. Experiments results show that compared with the OCM，the algorithm can improve the recall rate by 6.5%. When the recall rate is close，the algorithm reduce the amount of candidate boxes by 41.9%.
Keywords：texture-less object；object detection；template matching；object candidate box extraction；quantized and encoded orientation；binary gradient direction compression
参考文献：
[1] REDMON J，ANGELOVA A. Real-Time Grasp DetectionUsing Convolutional Neural Networks [J].Proceedings-IEEE International Conference on Robotics and Automation 2015，2014：1316-1322.
[2] DATTA R，JOSHI D，LI J，et al. Image Retrieval：Ideas，Influences，and Trends of the New Age [J].ACM Trans. on Computing Surveys，2008：40（2）：5.
[3] HODAN T，DAMEN D，WALTERIO M，et al. EFFICIENT TEXTURE-LESS OBJECT DETECTION FOR AUGMENTED REALITY GUIDANCE [J].2015 IEEE International Symposium on Mixed and Augment Reality Workshops，2015：81-86.
[4] LOWE D G. Distinctive Image Features from Scale-Invariant Keypoints [J].International Journal of Computer Vision，2004，60（2）：91-110.
[5] TOMBARI F，FRANCHI A，DI L. BOLD Features to Detect Texture-less Objects [C]//2013 IEEE International Conference on Computer Vision （ICCV）. IEEE Computer Society，2013：1265-1272.
[6] CHAN J，LEE J A，KEMAO Q. BORDER：An Oriented Rectangles Approach to Texture-Less Object Recognition [C]//Computer Vision & Pattern Recognition. IEEE，2016：2855-2863.
[7] HSIAO E，HEBERT M. Gradient Networks：Explicit Shape Matching Without Extracting Edges [C]//AAAI’13：Proceedings of the Twenty-Seventh AAAI Conference on Artificial Intelligence，2013：417-423.
[8] HINTERSTOISSER S，CAGNIART C，ILIC S，et al.Gradient Response Maps for Real-Time Detection of Texture-Less Objects [J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2012，1（5）：876-888.
[9] YU H，QIN H，PENG M. A Fast Approach to Texture-Less Object Detection Based on Orientation Compressing Map and Discriminative Regional Weight [J].Algorithms，2018，11（12）：201.
作者简介：彭茂庭（1995-），男，汉族，湖南邵阳人，硕士研究生，研究方向：数字图像处理、机器视觉。