Image semantic segmentation plays an important role and has application value in robot arm object capture, automatic driving, medical image analysis, geographic information system, etc. Aiming at the semantic segmentation method of deep learning(DL), this paper makes some attempts in the direction of weak supervised semantic segmentation, proposes AI and DL technology, and applies them to the Image segmentation technology(IST) in the construction machinery(CM) grasping task for analysis and exploration. The improvement of image weak location based on depth learning, network segmentation graph and image classification objective function are briefly analyzed. Finally, the experimental test analysis shows that a better initial attention map is not only helpful for direct segmentation, but also can promote the joint optimization performance, which verifies the effectiveness and feasibility of IST in the CM grasping task relying on AI and depth learning in this paper.

Umma Kavita. Visual Intelligent Recognition System based on Visual Thinking. Kinetic Mechanical Engineering (2021), Vol. 2, Issue 1: 46-54. https://doi.org/10.38007/KME.2021.020106.

[1] Rusyn B P , Lutsyk O A , Kosarevych R Y , et al. Image segmentation of clouds based on deep learning. Information Extraction and Processing, 2020, 2020(48):72-78. https://doi.org/10.15407/vidbir2020.48.072

[2] Li M X , Yu S Q , Zhang W , et al. Segmentation of retinal fluid based on deep learning: application of three-dimensional fully convolutional neural networks in optical coherence tomography images. International Journal of Ophthalmology, 2019, 12(6):1012-1020. https://doi.org/10.18240/ijo.2019.06.22

[3] Utomo T W , Cahyadi A I , Ardiyanto I . Suction-based Grasp Point Estimation in Cluttered Environment for Robotic Manipulator Using Deep Learning-based Affordance Map. International Journal of Automation and Computing, 2021, 18(2):277-287. https://doi.org/10.1007/s11633-020-1260-1

[4] Lee K , Kim J H , Lee H , et al. Boundary-Oriented Binary Building Segmentation Model With Two Scheme Learning for Aerial Images. IEEE Transactions on Geoscience and Remote Sensing, 2021, PP(99):1-17. https://doi.org/10.1109/TGRS.2021.3089623

[5] Brutti F , Fantazzini A , Finotello A , et al. Deep Learning to Automatically Segment and Analyze Abdominal Aortic Aneurysm from Computed Tomography Angiography. Cardiovascular Engineering and Technology, 2021, 13(4):535-547. https://doi.org/10.1007/s13239-021-00594-z

[6] Park J , Kwon D , Choi B W , et al. Small object segmentation with fully convolutional network based on overlapping domain decomposition. Machine Vision and Applications, 2019, 30(4):1-10. https://doi.org/10.1007/s00138-019-01023-x

[7] Imagawa H , Iwata M , Kise K . Digital Watermarking Method for Printed Matters Using Deep Learning for Detecting Watermarked Areas. IEICE Transactions on Information and Systems, 2021, E104.D(1):34-42. https://doi.org/10.1587/transinf.2020MUP0004

[8] Tsai C Y , Chou Y S , Wong C C , et al. Visually Guided Picking Control of an Omnidirectional Mobile Manipulator Based on End-to-End Multi-Task Imitation Learning. IEEE Access, 2019, PP(99):1-1.

[9] Li J , Yin J , Deng L . A robot vision navigation method using deep learning in edge computing environment. EURASIP Journal on Advances in Signal Processing, 2021, 2021(1):1-20. https://doi.org/10.1186/s13634-021-00734-6

[10] Hariyani Y S , Eom H , Park C . DA-CapNet: Dual Attention Deep Learning based on U-Net for Nailfold Capillary Segmentation. IEEE Access, 2020, PP(99):1-1. https://doi.org/10.1109/ACCESS.2020.2965651

[11] Utomo T W , Cahyadi A I , Ardiyanto I . Suction-based Grasp Point Estimation in Cluttered Environment for Robotic Manipulator Using Deep Learning-based Affordance Map. International Journal of Automation and Computing, 2021, 18(2):277-287. https://doi.org/10.1007/s11633-020-1260-1

[12] Tavousi P , Shahbazi Z , Shahbazmohamadi S . AI-based Brain Image Segmentation Using Synthesized Data. Microscopy and Microanalysis, 2020, 26(S2):694-697. https://doi.org/10.1017/S143192762001555X

[13] Park I , Kim K , Choi G , et al. AI-based Automatic Spine CT Image Segmentation and Haptic Rendering for Spinal Needle Insertion Simulator. The Journal of Korea Robotics Society, 2020, 15(4):316-322. https://doi.org/10.7746/jkros.2020.15.4.316

[14] Gautam S , Singhai J . Cosine-similarity watershed algorithm for water-body segmentation applying deep neural network classifier. Environmental Earth Sciences, 2021, 81(9):1-16. https://doi.org/10.1007/s12665-022-10376-y

[15] Ahamad M , Maizul E . Digital Analysis of Geo-Referenced Concrete Scanning Electron Microscope (SEM) Images. Civil And Environmental Engineering Reports, 2020, 30(2):65-79. https://doi.org/10.2478/ceer-2020-0020

[16] Chetty G , Yamin M , White M . A low resource 3D U-Net based deep learning model for medical image analysis. International Journal of Information Technology, 2021, 14(1):95-103. https://doi.org/10.1007/s41870-021-00850-4

[17] Saravana K , Sivaprakasam S A , Rengasari N E , et al. Fast K-Means technique for hyper-spectral image segmentation by multiband reduction. Pollack Periodica, 2019, 14(3):201-212. https://doi.org/10.1556/606.2019.14.3.19

[18] Ghosh S , Pal A , Jaiswal S , et al. SegFast-V2: Semantic image segmentation with less parameters in deep learning for autonomous driving. International journal of machine learning and cybernetics, 2019, 10(11):3145-3154. https://doi.org/10.1007/s13042-019-01005-5