Defaults to move the images into clusters folder

.gitignore

@@ -0,0 +1,5 @@
+.directory
+*.pyc
+__pycache__/
+data/
+clusters/

README.md

@@ -0,0 +1,21 @@
+super simple example of image clustering with image hashes ([wavelet hashing](https://fullstackml.com/wavelet-image-hash-in-python-3504fdd282b5) in particular) and DBSCAN.
+
+install deps: `pip3 install -r requirements.txt`
+
+to run:
+ - store all images in `static/img/` directory
+ - empty `data` directory if u wish to do a re-run of clustering
+ - run `python3 server.py --server` to do an interactive run and decide the cluster distance and min-samples count.
+ - run without `--server` argument to move the images to `clusters` directory
+
+```
+usage: server.py [-h] [-s] [--cluster distance] [--min-samples count]
+
+Image Cluster
+
+optional arguments:
+  -h, --help           show this help message and exit
+  -s, --server         run flask server (default: False)
+  --cluster distance   cluster distance (default: 1)
+  --min-samples count  minimum number of samples in a cluster (default: 1)
+```

cluster.py

@@ -1,3 +1,4 @@
+import os
 import imagehash
 import numpy as np
 from PIL import Image
@@ -38,4 +39,12 @@ def cluster(mat, fnames, eps, min_samples):
     clusters = defaultdict(list)
     for i, lbl in enumerate(labels):
         clusters[lbl].append(fnames[i])
-    return clusters
+    return clusters
+
+def create_directory(directory):
+    try:
+        os.makedirs(directory)
+        return True
+    except OSError:
+        if not os.path.isdir(directory): return False
+        return True

requirements.txt

@@ -1,3 +1,4 @@
-ImageHash==3.4
-scikit-learn==0.18
-numpy==1.11.1
+ImageHash
+scikit-learn
+numpy
+flask

server.py

@@ -1,16 +1,38 @@
 import json
+import shutil
 import numpy as np
 from glob import glob
-from flask import Flask, request, render_template, abort
-from cluster import compute_hashes, compute_dists, cluster
+from cluster import compute_hashes, compute_dists, cluster, create_directory
+
+import argparse
+parser = argparse.ArgumentParser(description='Image Cluster')
+
+# begin arguments
+parser.add_argument('-s', '--server', action='store_true',
+                    help='run flask server (default: %(default)s)\n\n')
+parser.add_argument('--cluster', metavar='distance', default=1,
+                    type=float, help='cluster distance (default: %(default)s)')
+parser.add_argument('--min-samples', metavar='count', default=1,
+                    type=int, help='minimum number of samples in a cluster (default: %(default)s)')
+
+args = parser.parse_args()
 
 
 if __name__ == '__main__':
+
+    if not create_directory('static/img/'):
+        print('Couldn\'t create static images directory!')
+        raise SystemExit
+
     try:
         mat = np.load('data/dist_mat.npy')
         fnames = json.load(open('data/fnames.json', 'r'))
     except FileNotFoundError:
-        print('computing hashes & distance matrix...')
+        if not create_directory('data'):
+            print('Couldn\'t create data directory!')
+            raise SystemExit
+
+        print('Computing hashes & distance matrix...')
         hashes, fnames = compute_hashes(glob('static/img/*'))
         mat = compute_dists(hashes)
 
@@ -19,25 +41,41 @@
             json.dump(fnames, f)
         np.save('data/dist_mat.npy', mat)
 
-    clusters = {}
-    app = Flask(__name__)
-
-    @app.route('/', methods=['GET', 'POST'])
-    def index():
-        global clusters
-        if request.method == 'POST':
-            eps = float(request.form.get('eps', 20))
-            min_samples = int(request.form.get('min_samples', 2))
-            clusters = cluster(mat, fnames, eps, min_samples)
-            print(clusters)
-        return render_template('index.html', clusters=clusters)
-
-    @app.route('/cluster/<int:id>', methods=['GET', 'POST'])
-    def view_cluster(id):
-        try:
-            print(clusters[id])
-            return render_template('cluster.html', cluster=clusters[id])
-        except KeyError:
-            abort(404)
-
-    app.run(host='0.0.0.0', port=5001)
+
+    if not args.server:
+        eps = args.cluster
+        min_samples = args.min_samples
+        clusters = cluster(mat, fnames, eps, min_samples)
+
+        print('Generated {} clusters!'.format(len(clusters)))
+        print('Moving images to "clusters" directory..')
+        for cluster in clusters:
+            create_directory('clusters/{}'.format(cluster))
+            for image in clusters[cluster]:
+                shutil.move(image, 'clusters/{}/'.format(cluster))
+        print('Created {} clusters and stored the images in "clusters" directory!'.format(len(clusters)))
+
+    else:
+        from flask import Flask, request, render_template, abort
+        clusters = {}
+        app = Flask(__name__)
+
+        @app.route('/', methods=['GET', 'POST'])
+        def index():
+            global clusters
+            if request.method == 'POST':
+                eps = float(request.form.get('eps', 20))
+                min_samples = int(request.form.get('min_samples', 2))
+                clusters = cluster(mat, fnames, eps, min_samples)
+                print(clusters)
+            return render_template('index.html', clusters=clusters)
+
+        @app.route('/cluster/<int:id>', methods=['GET', 'POST'])
+        def view_cluster(id):
+            try:
+                print(clusters[id])
+                return render_template('cluster.html', cluster=clusters[id])
+            except KeyError:
+                abort(404)
+
+        app.run(host='0.0.0.0', port=5002)