
Automatic Measurement of Multi-Posture Beef Cattle Body Size Based on Depth Image
YE Wenshuai, KANG Xi, HE Zhijiang, LI Mengfei, LIU Gang
Automatic Measurement of Multi-Posture Beef Cattle Body Size Based on Depth Image
Beef cattle in the farm are active, which leads the collection of posture of the beef cattle changeable, so it is difficult to automatically measure the body size of the beef cattle. Aiming at the above problems, an automatic measurement method for beef cattle's body size under multi-pose was proposed by analyzing the skeleton features of beef cattle head and the edge contour features of beef cattle images. Firstly, the consumer-grade depth camera Azure Kinect DK was used to collect the top-view depth video data directly above the beef cattle and the video data were divided into frames to obtain the original depth image. Secondly, the original depth image was processed by shadow interpolation, normalization, image segmentation and connected domain to remove the complex background and obtain the target image containing only beef cattle. Thirdly, the Zhang-Suen algorithm was used to extract the beef cattle skeleton of the target image, and calculated the intersection points and endpoints of the skeleton, so as to analyze the characteristics of the beef cattle head to determine the head removal point, and to remove the beef cattle head information from the image. Finally, the curvature curve of the beef cattle profile was obtained by the improved U-chord curvature method. The body measurement points were determined according to the curvature value and converted into three-dimensional spaces to calculate the body size parameters. In this paper, the postures of beef cattle, which were analyzed by a large amount of depth image data, were divided into left crooked, right crooked, correct posture, head down and head up, respectively. The test results showed that the head removal method proposed based on the skeleton in multiple postures hads head removel success rate higher than 92% in the five postures. Using the body measurement point extraction method based on the improved U-chord curvature proposed, the average absolute error of body length measurement was 2.73 cm, the average absolute error of body height measurement was 2.07 cm, and the average absolute error of belly width measurement was 1.47 cm. The method provides a better way to achieve the automatic measurement of beef cattle body size in multiple poses.
beef body size measurement / depth image / multi-gesture / body size measurement / Zhang-Suen algorithm / improved U-chord curvature body {{custom_keyword}} /
Table 1 Head removal results of five postures表1 五种姿态下肉牛头部去除结果 |
肉牛姿态图像 | 姿态端正图像 | 左歪图像 | 右歪图像 | 低头图像 | 抬头图像 |
---|---|---|---|---|---|
总帧数/帧 | 298 | 194 | 182 | 252 | 30 |
凸包分析方法成功帧数/帧 | 126 | 90 | 36 | 90 | 18 |
本文方法成功帧数/帧 | 282 | 186 | 168 | 235 | 30 |
凸包分析方法成功率/% | 42.28 | 46.39 | 19.78 | 35.71 | 60 |
本文方法成功率/% | 94.63 | 98.88 | 92.31 | 93.25 | 100 |
Table 2 Head removal results after the left-slanted posture was subdivided表 2 左歪姿态细分后肉牛头部去除结果 |
肉牛姿态 | 总帧 数/帧 | 凸包分析方法成功帧数/帧 | 本研究方法成功帧数/帧 | |
---|---|---|---|---|
轻度左歪图像 | 100 | 62 | 98 | 21.75 |
重度左歪图像 | 94 | 28 | 88 | 47.46 |
Table 3 Head removal result after the right-slanted posture was subdivided表 3 右歪姿态细分后肉牛头部去除结果 |
肉牛姿态 | 总帧 数/帧 | 凸包分析方法成功帧数/帧 | 本研究方法成功帧数/帧 | |
---|---|---|---|---|
轻度右歪图像 | 84 | 24 | 82 | 22.09 |
重度右歪图像 | 98 | 10 | 86 | 42.56 |
Table 4 Measuring the body size parameters of beef cattle by improved U-chord length curvature algorithm表 4 改进U弦长曲率算法测量肉牛体尺参数结果 |
肉牛编号 | 本文方法测量值 | 人工测量值 | 姿态 | ||||
---|---|---|---|---|---|---|---|
体高 | 腹宽 | 体直长 | 体高 | 腹宽 | 体直长/m | ||
1 | 1.2803 | 0.5991 | 1.2922 | 1.2651 | 0.5806 | 1.3146 | 低头 |
2 | 1.2784 | 0.5968 | 1.2577 | 1.2651 | 0.5806 | 1.3146 | 右歪 |
3 | 1.2528 | 0.5661 | 1.2546 | 1.2678 | 0.5790 | 1.2615 | 低头 |
4 | 1.2984 | 0.5968 | 1.2577 | 1.2678 | 0.5790 | 1.2615 | 右歪 |
5 | 1.2941 | 0.5852 | 1.2547 | 1.3121 | 0.5521 | 1.2520 | 右歪 |
6 | 1.3226 | 0.5761 | 1.2472 | 1.3121 | 0.5521 | 1.2520 | 姿态端正 |
7 | 1.3161 | 0.5531 | 1.2726 | 1.2946 | 0.5440 | 1.2260 | 姿态端正 |
8 | 1.3283 | 0.5595 | 1.2613 | 1.2946 | 0.5440 | 1.2260 | 左歪 |
9 | 1.2455 | 0.6524 | 1.3217 | 1.2259 | 0.6522 | 1.3523 | 低头 |
10 | 1.2392 | 0.6706 | 1.3115 | 1.2259 | 0.6522 | 1.3523 | 左歪 |
11 | 1.1279 | 0.6435 | 1.3749 | 1.2258 | 0.6486 | 1.3192 | 姿态端正 |
12 | 1.2420 | 0.6542 | 1.3389 | 1.2258 | 0.6486 | 1.3192 | 左歪 |
13 | 1.2813 | 0.5836 | 1.2308 | 1.2793 | 0.5575 | 1.3111 | 低头 |
14 | 1.3032 | 0.5765 | 1.2477 | 1.2893 | 0.5575 | 1.3111 | 左歪 |
15 | 1.2773 | 0.5829 | 1.1892 | 1.2348 | 0.6002 | 1.1899 | 右歪 |
16 | 1.2645 | 0.5904 | 1.1791 | 1.2348 | 0.6002 | 1.1899 | 姿态端正 |
17 | 1.2872 | 0.5736 | 1.1327 | 1.2606 | 0.5909 | 1.1365 | 右歪 |
18 | 1.2767 | 0.6093 | 1.1140 | 1.2606 | 0.5909 | 1.1365 | 低头 |
19 | 1.2565 | 0.6164 | 1.1514 | 1.2348 | 0.6002 | 1.1899 | 右歪 |
20 | 1.2422 | 0.6187 | 1.1553 | 1.2348 | 0.6002 | 1.1899 | 右歪 |
21 | 1.3331 | 0.5295 | 1.2316 | 1.3187 | 0.5189 | 1.2039 | 姿态端正 |
22 | 1.3414 | 0.5291 | 1.2271 | 1.3187 | 0.5189 | 1.2039 | 右歪 |
23 | 1.2987 | 0.5972 | 1.2444 | 1.2554 | 0.6207 | 1.2914 | 右歪 |
24 | 1.2768 | 0.6066 | 1.2731 | 1.2554 | 0.6207 | 1.2914 | 低头 |
25 | 1.2527 | 0.6408 | 1.2174 | 1.2635 | 0.6209 | 1.2267 | 低头 |
26 | 1.2793 | 0.6392 | 1.1921 | 1.2635 | 0.6209 | 1.2267 | 右歪 |
27 | 1.2690 | 0.6048 | 1.2176 | 1.2506 | 0.6144 | 1.2206 | 左歪 |
28 | 1.2631 | 0.5996 | 1.2079 | 1.2506 | 0.6144 | 1.2206 | 姿态端正 |
29 | 1.3248 | 0.5848 | 1.2481 | 1.3444 | 0.5956 | 1.2821 | 端正 |
30 | 1.3271 | 0.6140 | 1.3003 | 1.3444 | 0.5956 | 1.2821 | 低头 |
31 | 1.2858 | 0.6969 | 1.1688 | 1.2549 | 0.6986 | 1.2005 | 右歪 |
32 | 1.2360 | 0.6890 | 1.1303 | 1.2549 | 0.6986 | 1.2005 | 右歪 |
33 | 1.2803 | 0.6457 | 1.2163 | 1.2631 | 0.6465 | 1.2443 | 姿态端正 |
34 | 1.2509 | 0.6486 | 1.2287 | 1.2631 | 0.6465 | 1.2443 | 左歪 |
35 | 1.3167 | 0.5935 | 1.3026 | 1.3263 | 0.5825 | 1.3017 | 低头 |
36 | 1.3260 | 0.5947 | 1.3298 | 1.3263 | 0.5825 | 1.3017 | 端正 |
37 | 1.2250 | 0.6359 | 1.2869 | 1.2403 | 0.6508 | 1.2896 | 低头 |
38 | 1.1959 | 0.6546 | 1.2919 | 1.2403 | 0.6508 | 1.2896 | 端正 |
39 | 1.2190 | 0.5985 | 1.2307 | 1.2321 | 0.6069 | 1.2356 | 右歪 |
40 | 1.1941 | 0.5960 | 1.2789 | 1.2321 | 0.6069 | 1.2356 | 端正 |
41 | 1.3209 | 0.6147 | 1.1853 | 1.3187 | 0.6285 | 1.2179 | 右歪 |
42 | 1.3093 | 0.6183 | 1.1557 | 1.3187 | 0.6285 | 1.2179 | 右歪 |
43 | 1.2517 | 0.6017 | 1.1886 | 1.2674 | 0.6176 | 1.2035 | 姿态端正 |
44 | 1.2472 | 0.5744 | 1.1662 | 1.2674 | 0.6176 | 1.2035 | 右歪 |
45 | 1.4028 | 0.5970 | 1.2225 | 1.3737 | 0.6276 | 1.2762 | 右歪 |
46 | 1.3898 | 0.6091 | 1.2561 | 1.3737 | 0.6276 | 1.2762 | 右歪 |
Table 5 Average errors of body measure measurement under different postures表 5 不同姿态下肉牛体尺测量平均误差 |
肉牛姿态 | 体高平均 误差/cm | 体直长平均 误差/cm | 腹宽平均 误差/cm |
---|---|---|---|
端正姿态 | 2.742 | 2.515 | 1.126 |
非端正姿态 | 1.932 | 2.779 | 1.544 |
1 |
Food and Agriculture Organization of the United Nations, FAOSTAT[EB/OL].[2022-02-17].
{{custom_citation.content}}
{{custom_citation.annotation}}
|
2 |
李孟飞, 康熙, 王彦超, 等. 基于阈值分析的曲线拟合和边缘检测生猪侧视点云补全[J]. 农业机械学报, 2021, 52(S1): 81-88.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
3 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
4 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
5 |
NIR O,
{{custom_citation.content}}
{{custom_citation.annotation}}
|
6 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
7 |
赵建敏, 赵成, 夏海光. 基于Kinect v4的牛体尺测量方法[J]. 计算机应用, 2022, 42(5): 1598-1606.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
8 |
李嘉位, 马为红, 李奇峰, 等. 复杂环境下肉牛三维点云重建与目标提取方法[J]. 智慧农业(中英文), 2022, 4(2): 64-76.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
9 |
赵新强. 基于Kinect的奶牛体尺检测与试验研究[D]. 泰安: 山东农业大学, 2018.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
10 |
常海天. 种牛体尺非接触测量系统研究[D]. 长春:长春工业大学, 2018.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
11 |
司永胜, 安露露, 刘刚, 等. 基于Kinect相机的猪体理想姿态检测与体尺测量[J]. 农业机械学报, 2019, 50(1): 58-65.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
12 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
13 |
张文飞, 韩建海, 郭冰菁, 等. 改进的分水岭算法在粘连图像分割中的应用[J]. 计算机应用与软件, 2021, 38(6): 243-248.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
14 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
15 |
郭娟娟. 基于U弦长曲率的碎片拼接算法[D]. 苏州: 苏州大学, 2014.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
16 |
牛金玉. 基于三维点云的奶牛体尺测量与体重预测方法研究[D]. 杨凌 :西北农林科技大学, 2018.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
17 |
张馨月. 基于深度相机的生猪体尺自动测量技术研究[D]. 北京: 中国农业大学, 2021
{{custom_citation.content}}
{{custom_citation.annotation}}
|
18 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
19 |
李琦, 刘伟, 赵建敏. 基于双目视觉及Mask RCNN的牛体尺无接触测量[J]. 黑龙江畜牧兽医, 2020(6): 46-50, 159-160.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
{{custom_ref.label}} |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
Collection(s)
/
〈 |
|
〉 |