improve image hashing

atlas/models.py

@@ -892,7 +892,7 @@ class UncategorisedDicom(models.Model):
 
     series_instance_uid = models.CharField(max_length=255, blank=True, null=True)
 
-    image_md5_hash = models.CharField(max_length=32, null=True, blank=True)
+    image_blake3_hash = models.CharField(max_length=64, null=True, blank=True)
 
     created_date = models.DateTimeField(default=timezone.now)
 
@@ -901,12 +901,13 @@ class UncategorisedDicom(models.Model):
     def check_for_duplicates(self, image_hash):
         duplicate = None
 
-        if obj := UncategorisedDicom.objects.filter(image_md5_hash=image_hash).first():
-            duplicate = obj
+        if obj := UncategorisedDicom.objects.filter(image_blake3_hash=image_hash).first():
+            #duplicate = obj
+            return obj
 
         print(image_hash)
 
-        if obj := SeriesImage.objects.filter(image_md5_hash=image_hash).first():
+        if obj := SeriesImage.objects.filter(image_blake3_hash=image_hash).first():
             duplicate = obj
 
         return duplicate
@@ -920,11 +921,11 @@ class UncategorisedDicom(models.Model):
 
             # if self.check_for_duplicates():
             #    return DuplicateDicom
-            image_hash = get_image_dicom_hash(None, dataset=ds)
+            image_blake3_hash = get_image_dicom_hash(None, dataset=ds, hash_type="blake3", direct_pixel_data=True)
 
-            self.image_md5_hash = image_hash
+            self.image_blake3_hash = image_blake3_hash
 
-            duplicate = self.check_for_duplicates(image_hash)
+            duplicate = self.check_for_duplicates(image_blake3_hash)
 
             self.basic_dicom_tags = self.get_basic_dicom_tags(dataset=ds)
 
@@ -932,10 +933,8 @@ class UncategorisedDicom(models.Model):
                 if duplicate != self:
                     raise DuplicateDicom
 
-            
-
         # Hack for tests
-        if image_hash != "1234":
+        if image_blake3_hash != "1234":
             super().save(*args, **kwargs)  # Call the "real" save() method.
 
     def get_dicom_info(self):