你可以基于landmark这4个中心点，写个算法，将周围直径内的所有点从landmark72中穷举出所有相交点，这些点就是对应眼耳鼻的所有细节点

太阳穴应该是没有的。他抓了4个点是 双眼 耳 鼻

你看到下面有个landmark不。里头头2个就是左右眼

第几个是 左右眼中心点数据, 还有左右太阳穴的数据?  有具体文档吗

结果是返回来了,我要怎么把这些二维坐标点 带回到 图片jpg中比较好, 求推荐, 而且我只想要左右眼的中心点的数据,,

现在一堆的x  y数据过来,  我应该怎么使用??

{
"result": [{
"landmark72": [
{
"x": 77.646629333496,
"y": 303.31692504883
},
{
"x": 86.079063415527,
"y": 366.85278320312
},
{
"x": 99.066627502441,
"y": 430.21200561523
},
{
"x": 117.73243713379,
"y": 492.68936157227
},
{
"x": 165.6840057373,
"y": 553.28540039062
},
{
"x": 231.48829650879,
"y": 595.72814941406
},
{
"x": 295.59796142578,
"y": 606.23010253906
},
{
"x": 350.68078613281,
"y": 582.62231445312
},
{
"x": 397.41687011719,
"y": 530.57360839844
},
{
"x": 427.31915283203,
"y": 474.53518676758
},
{
"x": 442.0876159668,
"y": 417.94671630859
},
{
"x": 453.35983276367,
"y": 361.84887695312
},
{
"x": 458.65209960938,
"y": 307.32788085938
},
{
"x": 168.52355957031,
"y": 326.31860351562
},
{
"x": 189.06553649902,
"y": 313.85162353516
},
{
"x": 211.74104309082,
"y": 311.74780273438
},
{
"x": 232.83911132812,
"y": 318.29150390625
},
{
"x": 249.64990234375,
"y": 336.96490478516
},
{
"x": 230.92413330078,
"y": 342.00811767578
},
{
"x": 208.19184875488,
"y": 344.31896972656
},
{
"x": 185.69348144531,
"y": 338.49194335938
},
{
"x": 207.53587341309,
"y": 325.30059814453
},
{
"x": 137.64337158203,
"y": 284.50689697266
},
{
"x": 172.68797302246,
"y": 262.69598388672
},
{
"x": 208.3196105957,
"y": 265.98950195312
},
{
"x": 239.69641113281,
"y": 275.04281616211
},
{
"x": 268.74078369141,
"y": 297.7001953125
},
{
"x": 236.42715454102,
"y": 292.87658691406
},
{
"x": 205.28300476074,
"y": 285.66143798828
},
{
"x": 171.7755279541,
"y": 281.44018554688
},
{
"x": 341.59545898438,
"y": 337.34918212891
},
{
"x": 358.97100830078,
"y": 317.97454833984
},
{
"x": 380.10665893555,
"y": 311.81622314453
},
{
"x": 400.34533691406,
"y": 314.28921508789
},
{
"x": 417.01889038086,
"y": 325.84219360352
},
{
"x": 403.61267089844,
"y": 337.18103027344
},
{
"x": 383.40707397461,
"y": 342.92053222656
},
{
"x": 361.0803527832,
"y": 341.12194824219
},
{
"x": 375.37844848633,
"y": 325.24993896484
},
{
"x": 333.53787231445,
"y": 299.83056640625
},
{
"x": 359.23812866211,
"y": 278.03662109375
},
{
"x": 387.63900756836,
"y": 267.98590087891
},
{
"x": 417.87521362305,
"y": 264.31155395508
},
{
"x": 444.47152709961,
"y": 285.42236328125
},
{
"x": 419.26669311523,
"y": 282.90045166016
},
{
"x": 390.6819152832,
"y": 287.03637695312
},
{
"x": 362.77038574219,
"y": 294.53149414062
},
{
"x": 275.77352905273,
"y": 338.41204833984
},
{
"x": 270.00003051758,
"y": 370.64416503906
},
{
"x": 264.50018310547,
"y": 402.37185668945
},
{
"x": 250.8974609375,
"y": 435.87942504883
},
{
"x": 276.96731567383,
"y": 440.0549621582
},
{
"x": 327.19387817383,
"y": 440.77133178711
},
{
"x": 345.25219726562,
"y": 435.23751831055
},
{
"x": 334.44116210938,
"y": 402.67337036133
},
{
"x": 328.61529541016,
"y": 370.32012939453
},
{
"x": 322.41888427734,
"y": 339.15264892578
},
{
"x": 306.73321533203,
"y": 424.5087890625
},
{
"x": 228.8226776123,
"y": 498.28884887695
},
{
"x": 264.33874511719,
"y": 482.24429321289
},
{
"x": 301.40896606445,
"y": 480.71267700195
},
{
"x": 332.72491455078,
"y": 481.08358764648
},
{
"x": 357.25109863281,
"y": 495.84851074219
},
{
"x": 335.15069580078,
"y": 524.74206542969
},
{
"x": 299.18453979492,
"y": 538.18865966797
},
{
"x": 257.64117431641,
"y": 527.66149902344
},
{
"x": 266.20782470703,
"y": 496.01083374023
},
{
"x": 300.72848510742,
"y": 497.23867797852
},
{
"x": 330.85455322266,
"y": 494.94967651367
},
{
"x": 327.7646484375,
"y": 505.01931762695
},
{
"x": 298.05319213867,
"y": 509.83383178711
},
{
"x": 264.90447998047,
"y": 506.86260986328
}
],
"roll": -0.57837802171707,
"location": {
"top": 263,
"left": 78,
"width": 380,
"height": 343
},
"face_probability": 1,
"rotation_angle": 0,
"pitch": 11.926297187805,
"landmark": [
{
"x": 207.53587341309,
"y": 325.30059814453
},
{
"x": 375.37844848633,
"y": 325.24993896484
},
{
"x": 306.73321533203,
"y": 424.5087890625
},
{
"x": 297.50259399414,
"y": 502.76327514648
}
],
"yaw": -5.9680991172791
}],
"log_id": 2261232835122011,
"result_num": 1
}

人脸检测的时候加上landmark的选项，然后返回json中会多出landmark和landmark72.landmark表示眼口鼻。landmark72表示全脸细节点

具体的左右的位置数据应该怎么取??

"right_eye": 0,
"nose": 0,
"mouth": 0,
"chin": 0,
"left_eye": 0,  没有被遮挡???  我要具体的  右眼位置坐标和鼻尖的坐标,应该怎么写呀??

occlusion表示的是人脸各部分遮挡的概率,[0, 1],0表示完整，1表示不完整

0是说明没有被遮挡。是合法的值

翻一下文档，里头都有详细的说明