{"id":7652,"date":"2019-08-27T11:10:31","date_gmt":"2019-08-27T02:10:31","guid":{"rendered":"https:\/\/itport.cloud\/?p=7652"},"modified":"2019-08-27T11:10:31","modified_gmt":"2019-08-27T02:10:31","slug":"post-7652","status":"publish","type":"post","link":"https:\/\/itport.cloud\/?p=7652","title":{"rendered":"(\u7b2c5\u56de)Python + OpenCV \u3067\u904a\u3093\u3067\u307f\u308b(YOLO\u3092\u7528\u3044\u305f\u7269\u4f53\u691c\u51fa\u7de8)"},"content":{"rendered":"\n<h2 class=\"wp-block-heading\" id=\"e79baee6aca1-1\"><strong>\u76ee\u6b21<\/strong><\/h2>\n\n\n\n<ul class=\"wp-block-list\"><li><a href=\"#title-1\">\u306f\u3058\u3081\u306b<\/a><\/li><li><a href=\"#title-2\">YOLO\u3068\u306f<\/a><\/li><li><a href=\"#title-3\">\u5fc5\u8981\u306a\u30e9\u30a4\u30d6\u30e9\u30ea<\/a><\/li><li><a href=\"#title-4\"> YOLO\u3092\u7528\u3044\u305f\u7269\u4f53\u691c\u51fa<\/a><\/li><li><a href=\"#title-5\"> \u304a\u308f\u308a\u306b <\/a><\/li><\/ul>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"title-1\"><strong>\u306f\u3058\u3081\u306b<\/strong><\/h2>\n\n\n\n<p>\u524d\u56de\u307e\u3067\u306fOpenCV\u306b\u540c\u68b1\u3055\u308c\u3066\u3044\u308b\u30ab\u30b9\u30b1\u30fc\u30c9\u578b\u306e\u691c\u51fa\u5668\u3092\u7528\u3044\u3066\u3001\u9759\u6b62\u753b\u304a\u3088\u3073\u52d5\u753b\u3092\u4f7f\u3063\u3066\u9854\u691c\u51fa\u3092\u884c\u3044\u307e\u3057\u305f\u3002<br>\u4eca\u56de\u306f\u3001YOLO\u3068\u547c\u3070\u308c\u308b\u7269\u4f53\u691c\u51fa\u6cd5\u3092\u7528\u3044\u305f\u7269\u4f53\u691c\u51fa\u3092\u884c\u3063\u3066\u307f\u305f\u3044\u3068\u601d\u3044\u307e\u3059\u3002 <\/p>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"title-2\"><strong>YOLO\u3068\u306f<\/strong><\/h2>\n\n\n\n<p>YOLO\u3068\u306f\u300cYou Only Look Once\u300d\uff08\u4e00\u76ee\u898b\u308b\u3060\u3051\u3067\uff09\u306e\u982d\u6587\u5b57\u3092\u3068\u3063\u305f\u7565\u8a9e\u3067\u3001\u4e00\u76ee\u898b\u305f\u3060\u3051\u3067\u7269\u4f53\u3092\u691c\u51fa\u3067\u304d\u308b\u3068\u3044\u3046\u7279\u5fb4\u304c\u3042\u308a\u307e\u3059\u3002<br><br>\u524d\u56de\u307e\u3067\u306e\u691c\u51fa\u6cd5\u306f\u3001Sliding Window Approach\u306a\u624b\u6cd5\u3067\u3001\u753b\u50cf\u5185\u3092\u691c\u51fa\u7bc4\u56f2\u306e\u5927\u304d\u3055\u3092\u5909\u3048\u305f\u308a\u3001\u52d5\u304b\u3057\u305f\u308a\u3057\u3066\u8907\u6570\u56de\u306e\u691c\u8a3c\u3092\u884c\u3044\u3001\u30d1\u30bf\u30fc\u30f3\u306b\u4e00\u81f4\u3059\u308b\u90e8\u5206\u3092\u691c\u51fa\u3059\u308b\u624b\u6cd5\u3067\u3057\u305f\u3002<br><br>YOLO\u306f\u3001\u753b\u50cf\u3092\u4e00\u5ea6\u3060\u3051CNN(Convolutional Neural Network:\u7573\u307f\u8fbc\u307f\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af)\u306b\u901a\u3059\u3060\u3051\u3067\u3001\u753b\u50cf\u5185\u306e\u8907\u6570\u306e\u7269\u4f53\u3092\u691c\u51fa\u304a\u3088\u3073\u8a8d\u8b58\u3092\u540c\u6642\u306b\u884c\u3046\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u3067\u3059\u3002<br>1\u5ea6\u3060\u3051CNN\u3092\u901a\u3059\u3060\u3051\u306a\u306e\u3067\u3001\u9ad8\u901f\u5316\u304c\u56f3\u3089\u308c\u3066\u304a\u308a\u3001\u30ea\u30a2\u30eb\u30bf\u30a4\u30e0\u3067\u306e\u691c\u51fa\u30fb\u8a8d\u8b58\u306b\u3082\u3088\u304f\u7528\u3044\u3089\u308c\u3066\u3044\u307e\u3059\u3002<br><br>\u73fe\u5728\u306f2018\u5e74\u306b\u767a\u8868\u3055\u308c\u305fYOLOv3(Version 3)\u304c\u3042\u308a\u3001YOLOv2(Version 2)\u3067\u306fCNN\u304c19\u5c64\u30e2\u30c7\u30eb\u3067\u3042\u3063\u305f\u306e\u304cYOLOv3\u3067\u306f53\u5c64\u30e2\u30c7\u30eb\u306b\u306a\u3063\u3066\u304a\u308a\u3001\u8a8d\u8b58\u7cbe\u5ea6\u304c\u5927\u5e45\u306b\u30a2\u30c3\u30d7\u3057\u3066\u3044\u307e\u3059\u3002 <br><br><div class=\"linkcard\"><div class=\"lkc-external-wrap\"><a class=\"lkc-link no_icon\" href=\"https:\/\/pjreddie.com\/darknet\/yolo\/\" data-lkc-id=\"39\" target=\"_blank\" rel=\"external noopener\"><div class=\"lkc-card\"><div class=\"lkc-content\"><figure class=\"lkc-thumbnail\"><img decoding=\"async\" class=\"lkc-thumbnail-img\" src=\"https:\/\/s.wordpress.com\/mshots\/v1\/https%3A%2F%2Fpjreddie.com%2Fdarknet%2Fyolo%2F?w=100\" width=\"100px\" height=\"108px\" alt=\"\" \/><\/figure><div class=\"lkc-title\">YOLO: Real-Time Object Detection<\/div><div class=\"lkc-url\" title=\"https:\/\/pjreddie.com\/darknet\/yolo\/\">https:\/\/pjreddie.com\/darknet\/yolo\/<\/div><div class=\"lkc-excerpt\">You only look once (YOLO) is a state-of-the-art, real-time object detection system.<\/div><div class=\"lkc-more\">more<\/div><\/div><div class=\"lkc-info\"><div class=\"lkc-favicon\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/www.google.com\/s2\/favicons?domain=pjreddie.com\" alt=\"\" width=\"16\" height=\"16\" \/><\/div><div class=\"lkc-domain\">pjreddie.com<\/div><div class=\"lkc-share\"> <div class=\"lkc-sns-hb\">103 Users<\/div><\/div><\/div><div class=\"clear\"><\/div><\/div><\/a><\/div><\/div><\/p>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"title-3\"><strong>\u5fc5\u8981\u306a\u30e9\u30a4\u30d6\u30e9\u30ea<\/strong><\/h2>\n\n\n\n<p>\n\n\u4eca\u56de\u306fYOLOv3\u3067\u691c\u51fa\u30fb\u8a8d\u8b58\u3057\u3001OpenCV\u3092\u7528\u3044\u3066\u753b\u50cf\u51e6\u7406\u3057\u307e\u3059\u3002\n\n<\/p>\n\n\n\n<ul class=\"wp-block-list\"><li> YOLOv3:PyTorch\u7528\u306eYOLOv3\u3092\u7528\u3044\u307e\u3059\u3002 <\/li><\/ul>\n\n\n<div class=\"wp-block-syntaxhighlighter-code \"><pre class=\"brush: plain; title: ; notranslate\" title=\"\">\ngit clone https:\/\/github.com\/ayooshkathuria\/pytorch-yolo-v3.git\n<\/pre><\/div>\n\n\n<ul class=\"wp-block-list\"><li>PyTorch:Python\u306e\u6a5f\u68b0\u5b66\u7fd2\u30e9\u30a4\u30d6\u30e9\u30ea<\/li><li>Pandas:\u30c7\u30fc\u30bf\u89e3\u6790\u652f\u63f4\u30e9\u30a4\u30d6\u30e9\u30ea<\/li><li>Cython:Python\u3092C\/C++\u306b\u5909\u63db\u3059\u308b\u30e9\u30a4\u30d6\u30e9\u30ea\uff08\u9ad8\u901f\u5316\u3092\u56f3\u308b\u305f\u3081\uff09<\/li><\/ul>\n\n\n<div class=\"wp-block-syntaxhighlighter-code \"><pre class=\"brush: plain; title: ; notranslate\" title=\"\">\nconda install pandas\nconda install pytorch torchvision -c pytorch\npip install cython\n\nconda install opencv (OpenCV\u3092\u5c0e\u5165\u3057\u3066\u3044\u306a\u3044\u5834\u5408\u306e\u307f)\npip install matplotlib (matplotlib\u3092\u5c0e\u5165\u3057\u3066\u3044\u306a\u3044\u5834\u5408\u306e\u307f)\n<\/pre><\/div>\n\n\n<p>\n\n\u30e9\u30a4\u30d6\u30e9\u30ea\u3067\u306f\u306a\u3044\u3067\u3059\u304c\u3001\u91cd\u307f\u30d5\u30a1\u30a4\u30eb\u3092\u4ee5\u4e0b\u304b\u3089\u30c0\u30a6\u30f3\u30ed\u30fc\u30c9\u3057\u3066\u304a\u304d\u307e\u3059\u3002\n\n<\/p>\n\n\n<div class=\"wp-block-syntaxhighlighter-code \"><pre class=\"brush: plain; title: ; notranslate\" title=\"\">\ncd pytorch-yolo-v3\nwget https:\/\/pjreddie.com\/media\/files\/yolov3.weights\n<\/pre><\/div>\n\n\n<h2 class=\"wp-block-heading\" id=\"title-4\"><strong>YOLO\u3092\u7528\u3044\u305f\u7269\u4f53\u691c\u51fa<\/strong><\/h2>\n\n\n\n<p>\n\n\u4eca\u56de\u4f7f\u7528\u3059\u308b\u30d7\u30ed\u30b0\u30e9\u30e0\u306fPyTorch\u7528YOLOv3\u306b\u540c\u68b1\u3055\u308c\u3066\u3044\u308bvideo_demo.py\u306e\u5185\u5bb9\u3092\u6539\u5909\u3057\u305f\u3082\u306e\u3067\u3059\u3002\n\n<\/p>\n\n\n<div class=\"wp-block-syntaxhighlighter-code \"><pre class=\"brush: python; title: ; notranslate\" title=\"\">\nfrom __future__ import division\nimport time\nimport torch\nimport torch.nn as nn\nfrom torch.autograd import Variable\nimport numpy as np\nimport cv2\nfrom util import *\nfrom darknet import Darknet\nfrom preprocess import prep_image, inp_to_image\nimport pandas as pd\nimport random\nimport argparse\nimport pickle as pkl\n\ndef prep_image(img, inp_dim):\n    # CNN\u306b\u901a\u3059\u305f\u3081\u306b\u753b\u50cf\u3092\u52a0\u5de5\u3059\u308b\n    orig_im = img\n    dim = orig_im.shape&#x5B;1], orig_im.shape&#x5B;0]\n    img = cv2.resize(orig_im, (inp_dim, inp_dim))\n    img_ = img&#x5B;:,:,::-1].transpose((2,0,1)).copy()\n    img_ = torch.from_numpy(img_).float().div(255.0).unsqueeze(0)\n    return img_, orig_im, dim\n\ndef write(x, img):\n    # \u753b\u50cf\u306b\u7d50\u679c\u3092\u63cf\u753b\n    c1 = tuple(x&#x5B;1:3].int())\n    c2 = tuple(x&#x5B;3:5].int())\n    cls = int(x&#x5B;-1])\n    label = \"{0}\".format(classes&#x5B;cls])\n    color = random.choice(colors)\n    cv2.rectangle(img, c1, c2,color, 1)\n    t_size = cv2.getTextSize(label, cv2.FONT_HERSHEY_PLAIN, 1 , 1)&#x5B;0]\n    c2 = c1&#x5B;0] + t_size&#x5B;0] + 3, c1&#x5B;1] + t_size&#x5B;1] + 4\n    cv2.rectangle(img, c1, c2,color, -1)\n    cv2.putText(img, label, (c1&#x5B;0], c1&#x5B;1] + t_size&#x5B;1] + 4), cv2.FONT_HERSHEY_PLAIN, 1, &#x5B;225,255,255], 1);\n    return img\n\ndef arg_parse():\n    # \u30e2\u30b8\u30e5\u30fc\u30eb\u306e\u5f15\u6570\u3092\u4f5c\u6210\n    parser = argparse.ArgumentParser(description='YOLO v3 Cam Demo') # ArgumentParser\u3067\u5f15\u6570\u3092\u8a2d\u5b9a\u3059\u308b\n    parser.add_argument(\"--confidence\", dest = \"confidence\", help = \"Object Confidence to filter predictions\", default = 0.25)\n    # confidence\u306f\u4fe1\u983c\u6027\n    parser.add_argument(\"--nms_thresh\", dest = \"nms_thresh\", help = \"NMS Threshhold\", default = 0.4)\n    # nms_thresh\u306f\u95be\u5024\n   \n    parser.add_argument(\"--reso\", dest = 'reso', help =\n                        \"Input resolution of the network. Increase to increase accuracy. Decrease to increase speed\",\n                        default = \"160\", type = str)\n                        # reso\u306fCNN\u306e\u5165\u529b\u89e3\u50cf\u5ea6\u3067\u3001\u5897\u52a0\u3055\u305b\u308b\u3068\u7cbe\u5ea6\u304c\u4e0a\u304c\u308b\u304c\u3001\u901f\u5ea6\u304c\u4f4e\u4e0b\u3059\u308b\u3002\n    return parser.parse_args() # \u5f15\u6570\u3092\u89e3\u6790\u3057\u3001\u8fd4\u3059\n\nif __name__ == '__main__':\n    cfgfile = \"cfg\/yolov3.cfg\" # \u8a2d\u5b9a\u30d5\u30a1\u30a4\u30eb\n    weightsfile = \"yolov3.weights\" # \u91cd\u307f\u30d5\u30a1\u30a4\u30eb\n    num_classes = 80 # \u30af\u30e9\u30b9\u306e\u6570\n\n    args = arg_parse() # \u5f15\u6570\u3092\u53d6\u5f97\n    confidence = float(args.confidence) # \u4fe1\u983c\u6027\u306e\u8a2d\u5b9a\u5024\u3092\u53d6\u5f97\n    nms_thesh = float(args.nms_thresh) # \u95be\u5024\u3092\u53d6\u5f97\n    start = 0\n    CUDA = torch.cuda.is_available() # CUDA\u304c\u4f7f\u7528\u53ef\u80fd\u304b\u3069\u3046\u304b\n\n    num_classes = 80 # \u30af\u30e9\u30b9\u306e\u6570\n    bbox_attrs = 5 + num_classes\n\n    model = Darknet(cfgfile) #model\u306e\u4f5c\u6210\n    model.load_weights(weightsfile) # model\u306b\u91cd\u307f\u3092\u8aad\u307f\u8fbc\u3080\n\n    model.net_info&#x5B;\"height\"] = args.reso\n    inp_dim = int(model.net_info&#x5B;\"height\"])\n\n    assert inp_dim % 32 == 0\n    assert inp_dim > 32\n\n    if CUDA:\n        model.cuda() #CUDA\u304c\u4f7f\u7528\u53ef\u80fd\u3067\u3042\u308c\u3070cuda\u3092\u8d77\u52d5\n\n    model.eval()\n\n    cap = cv2.VideoCapture(0) #\u30ab\u30e1\u30e9\u3092\u6307\u5b9a\n\n    assert cap.isOpened(), 'Cannot capture source' #\u30ab\u30e1\u30e9\u304c\u8d77\u52d5\u3067\u304d\u305f\u304b\u78ba\u8a8d\n\n    frames = 0\n    start = time.time()\n    while cap.isOpened(): #\u30ab\u30e1\u30e9\u304c\u8d77\u52d5\u3057\u3066\u3044\u308b\u9593\n\n        ret, frame = cap.read() #\u30ad\u30e3\u30d7\u30c1\u30e3\u753b\u50cf\u3092\u53d6\u5f97\n        if ret:\n            # \u89e3\u6790\u6e96\u5099\u3068\u3057\u3066\u30ad\u30e3\u30d7\u30c1\u30e3\u753b\u50cf\u3092\u52a0\u5de5\n            img, orig_im, dim = prep_image(frame, inp_dim)\n\n            if CUDA:\n                im_dim = im_dim.cuda()\n                img = img.cuda()\n\n            output = model(Variable(img), CUDA)\n            output = write_results(output, confidence, num_classes, nms = True, nms_conf = nms_thesh)\n\n            # FPS\u306e\u8868\u793a\n            if type(output) == int:\n                frames += 1\n                print(\"FPS of the video is {:5.2f}\".format( frames \/ (time.time() - start)))\n                cv2.imshow(\"frame\", orig_im)\n\n                # q\u30ad\u30fc\u3092\u62bc\u3059\u3068FPS\u8868\u793a\u306e\u7d42\u4e86\n                key = cv2.waitKey(1)\n                if key & 0xFF == ord('q'):\n                    break\n                continue\n\n            output&#x5B;:,1:5] = torch.clamp(output&#x5B;:,1:5], 0.0, float(inp_dim))\/inp_dim\n            output&#x5B;:,&#x5B;1,3]] *= frame.shape&#x5B;1]\n            output&#x5B;:,&#x5B;2,4]] *= frame.shape&#x5B;0]\n\n            classes = load_classes('data\/coco.names') # \u8b58\u5225\u30af\u30e9\u30b9\u306e\u30ea\u30b9\u30c8\n            colors = pkl.load(open(\"pallete\", \"rb\"))\n\n            list(map(lambda x: write(x, orig_im), output))\n\n            cv2.imshow(\"frame\", orig_im)\n            key = cv2.waitKey(1)\n            # q\u30ad\u30fc\u3092\u62bc\u3059\u3068\u52d5\u753b\u8868\u793a\u306e\u7d42\u4e86\n            if key & 0xFF == ord('q'):\n                break\n            frames += 1\n            print(\"FPS of the video is {:5.2f}\".format( frames \/ (time.time() - start)))\n\n        else:\n            break\n<\/pre><\/div>\n\n\n<p>\n\n\u3053\u306e\u30d7\u30ed\u30b0\u30e9\u30e0\u3092Anaconda Prompt\u3092\u8d77\u52d5\u3057\u3066\u3001\u5b9f\u884c\u3057\u307e\u3059\u3002\n\n<\/p>\n\n\n<div class=\"wp-block-syntaxhighlighter-code \"><pre class=\"brush: plain; title: ; notranslate\" title=\"\">\npython \u4e0a\u8a18\u306e\u30d7\u30ed\u30b0\u30e9\u30e0\u3092\u5165\u529b\u3057\u305fpy\u30d5\u30a1\u30a4\u30eb.py\n<\/pre><\/div>\n\n\n<div class=\"wp-block-image\"><figure class=\"aligncenter\"><img loading=\"lazy\" decoding=\"async\" width=\"300\" height=\"225\" src=\"https:\/\/itport.cloud\/wp-content\/uploads\/2019\/08\/qsljbPNW.gif\" alt=\"\" class=\"wp-image-7655\"\/><figcaption>\u7269\u4f53\u691c\u51fa\u7d50\u679c<\/figcaption><\/figure><\/div>\n\n\n\n<p>\u30d6\u30ed\u30b0\u306b\u30a2\u30c3\u30d7\u30ed\u30fc\u30c9\u3059\u308b\u305f\u3081\u3001\u753b\u50cf\u304c\u7c97\u304f\u5c0f\u3055\u304f\u306a\u3063\u3066\u3057\u307e\u3044\u307e\u3057\u305f\u304c\u3001\u30da\u30c3\u30c8\u30dc\u30c8\u30eb\u304cbottle\uff08\u305f\u307e\u306bcup\uff09\u3001\u30c6\u30fc\u30d6\u30eb\u304cdining table\u3001\u6905\u5b50\u304cchair\u3068\u691c\u51fa\u30fb\u8a8d\u8b58\u3057\u3066\u3044\u307e\u3059\u3002<br>\u307e\u305f\u624b\u3092\u5dee\u3057\u51fa\u3059\u3068\u3001person\u3068\u3057\u3066\u624b\u304c\u4eba\u306e\u4e00\u90e8\u3060\u3068\u3082\u8a8d\u8b58\u3057\u3066\u3044\u307e\u3059\u3002<br><br>\u79c1\u306eWindows\u30bf\u30d6\u30ec\u30c3\u30c8\u3067\u306fFPS\u304c2\u7a0b\u5ea6\u3067\u3042\u308a\u3001\u591a\u5c11\u30ab\u30af\u30ab\u30af\u3057\u305f\u52d5\u304d\u306b\u306a\u308a\u307e\u3057\u305f\u3002<br>\u4eca\u56de\u306e\u51e6\u7406\u306fCPU\u306e\u307f\u3067\u52d5\u4f5c\u3057\u3066\u304a\u308a\u3001GPU\u3092\u642d\u8f09\u3057\u305f\u30de\u30b7\u30f3\u7b49\u3092\u4f7f\u7528\u3059\u308c\u3070\u3082\u3063\u3068\u30b9\u30e0\u30fc\u30ba\u306a\u52d5\u753b\u3067\u51e6\u7406\u304c\u884c\u3048\u308b\u304b\u3082\u3057\u308c\u307e\u305b\u3093\u306d\u3002 <\/p>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"title-5\"><strong>\u304a\u308f\u308a\u306b<\/strong><\/h2>\n\n\n\n<p> \u4eca\u56de\u306f\u3001YOLOv3\u3092\u5229\u7528\u3057\u3066\u7269\u4f53\u691c\u51fa\u3092\u884c\u3044\u307e\u3057\u305f\u3002<br>\u30ea\u30a2\u30eb\u30bf\u30a4\u30e0\u3067\u3082\u4f7f\u3048\u308b\u65e9\u3055\u3067\u7269\u4f53\u691c\u51fa\u304c\u51fa\u6765\u3066\u3044\u308b\u306e\u306f\u9a5a\u304d\u3067\u3057\u305f\u3002<br>\u307e\u305f\u3001\u7269\u4f53\u8a8d\u8b58\u306e\u7cbe\u5ea6\u304c\u9ad8\u304f\u3001\u7269\u4f53\u3092\u8a8d\u8b58\u3057\u3066\u3044\u308b\u69d8\u306f\u3001\u6620\u753b\u300c\u30bf\u30fc\u30df\u30cd\u30fc\u30bf\u30fc\u300d\u3092\u601d\u3044\u51fa\u3055\u305b\u307e\u3059\u306d\u3002<br><br>\u6b21\u56de\u306f\u3001\u753b\u50cf\u304b\u3089\u6587\u5b57\u8a8d\u8b58\u3092\u3055\u305b\u308bOCR\u3092\u884c\u3063\u3066\u307f\u305f\u3044\u3068\u601d\u3044\u307e\u3059\u3002 <\/p>\n","protected":false},"excerpt":{"rendered":"\u76ee\u6b21 \u306f\u3058\u3081\u306bYOLO\u3068\u306f\u5fc5\u8981\u306a\u30e9\u30a4\u30d6\u30e9\u30ea YOLO\u3092\u7528\u3044\u305f\u7269\u4f53\u691c\u51fa \u304a\u308f\u308a\u306b \u306f\u3058\u3081\u306b \u524d\u56de\u307e\u3067\u306fOpenCV\u306b\u540c\u68b1\u3055\u308c\u3066\u3044\u308b\u30ab\u30b9\u30b1\u30fc\u30c9\u578b\u306e\u691c\u51fa\u5668\u3092\u7528\u3044\u3066\u3001\u9759\u6b62\u753b\u304a\u3088\u3073\u52d5\u753b\u3092\u4f7f\u3063\u3066\u9854\u691c\u51fa\u3092\u884c\u3044\u307e\u3057\u305f\u3002\u4eca\u56de\u306f\u3001YOLO\u3068\u547c\u3070\u308c\u308b\u7269\u4f53\u691c\u51fa\u6cd5\u3092\u7528\u3044\u305f\u7269\u4f53\u691c\u51fa\u3092\u884c\u3063\u3066\u307f\u305f\u3044\u3068\u601d\u3044\u307e\u3059\u3002 YOLO\u3068\u306f YOLO\u3068\u306f\u300cYou Only Look Once\u300d\uff08\u4e00\u76ee\u898b\u308b\u3060\u3051\u3067\uff09\u306e\u982d\u6587\u5b57\u3092\u3068\u3063\u305f\u7565\u8a9e\u3067\u3001\u4e00\u76ee\u898b\u305f\u3060\u3051\u3067\u7269\u4f53\u3092\u691c\u51fa\u3067\u304d\u308b\u3068\u3044\u3046\u7279\u5fb4\u304c\u3042\u308a\u307e\u3059\u3002\u524d\u56de\u307e\u3067\u306e\u691c\u51fa\u6cd5\u306f\u3001Sliding Window Approach\u306a\u624b\u6cd5\u3067\u3001\u753b\u50cf\u5185\u3092\u691c\u51fa\u7bc4\u56f2\u306e\u5927\u304d\u3055\u3092\u5909\u3048\u305f\u308a\u3001\u52d5\u304b\u3057\u305f\u308a\u3057\u3066\u8907\u6570\u56de\u306e\u691c\u8a3c\u3092\u884c\u3044\u3001\u30d1\u30bf\u30fc\u30f3\u306b\u4e00\u81f4\u3059\u308b\u90e8\u5206\u3092\u691c\u51fa\u3059\u308b\u624b\u6cd5\u3067\u3057\u305f\u3002YOLO&hellip;","protected":false},"author":6,"featured_media":7774,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[51],"tags":[52],"class_list":{"0":"post-7652","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-opencv","8":"tag-season15"},"_links":{"self":[{"href":"https:\/\/itport.cloud\/index.php?rest_route=\/wp\/v2\/posts\/7652","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/itport.cloud\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/itport.cloud\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/itport.cloud\/index.php?rest_route=\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/itport.cloud\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=7652"}],"version-history":[{"count":14,"href":"https:\/\/itport.cloud\/index.php?rest_route=\/wp\/v2\/posts\/7652\/revisions"}],"predecessor-version":[{"id":7772,"href":"https:\/\/itport.cloud\/index.php?rest_route=\/wp\/v2\/posts\/7652\/revisions\/7772"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/itport.cloud\/index.php?rest_route=\/wp\/v2\/media\/7774"}],"wp:attachment":[{"href":"https:\/\/itport.cloud\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=7652"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/itport.cloud\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=7652"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/itport.cloud\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=7652"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}