Iskanje podvojenih video datotek

Zdravo.
Ali obstaja kak dober brezplačen program, ki bi poiskal dublirane video datoteke,
katere nimajo istega imena , iste resolucije, istega trajanja ali iste velikosti.
Najboljše, kar sem do sedaj našel je Duplicate Media Finder, ampak je plačljiv.
Najde ni da ni, ampak je v demo verziji večina funkcij onemogočena.

phnj90 je 3. feb 2023 ob 21:12 izjavil:

Hmm, zadeva JE zanimiva (meni osebno sicer niti ne), 45EUR je menda 1 full licence za na pet računalnikov... Poskušaj s 'crowd fundingom' na forumu, mogoče bo vzelo nekaj časa, ampak možnosti so.

Stand up komik ,ali ne....
Zakaj bi, preden ne izčrpam vseh možnosti, vključno z možnostjo pomoči tukaj, predčasno plačal ...
Ali ni forum namenjen ravno temu ,da si pomagamo, ne pa takoj začeti z nekimi bednimi nabijanji.
Mogoče bi ti rabil 'crowd funding' da bi se znebil demona.
Pa Lp.

Poskusi s tem
VisioForge Video Fingerprinting SDK
https://www.visioforge.com/video-finger...

še eden

Duplicate Video Search
https://duplicatevideosearch.com/how-to...


našel tu:
https://fixthephoto.com/best-duplicate-...
7 in 8




btw, lahko tudi vprašaš https://chat.openai.com/chat , da ti napiše npr. python kodo za iskanje podobnih posnetkov, z uporabo
OpenCV
https://pastebin.com/geAa83fU

ali
scikit-learn
https://pastebin.com/fNPzY1vB



Yes, it's possible to detect duplicate videos without using hash functions. One approach is to use machine learning to compare the visual content of the videos. Here's a high-level outline of the process:

1. Extract features from each video: This can be done by breaking down the video into individual frames and extracting features from each frame. Some examples of features that can be extracted include color histograms, edge detection, or local binary patterns.

2. Compare the features: Once you have extracted the features for each video, you can compare the features between all pairs of videos to see if they are similar. You can use a distance metric such as Euclidean distance or Cosine similarity to compare the features.

3. Identify duplicates: If the features of two videos are similar enough, you can consider them duplicates. You can set a threshold for the distance metric to determine what constitutes a duplicate.

Note that this approach is more complex and computationally expensive than using a hash function, but it has the advantage of being able to detect duplicates even if the videos have been slightly modified (e.g., by adding a watermark).

If you're interested in using machine learning to detect duplicate videos, you could use a pre-trained model, such as a deep neural network, or train your own model from scratch using a dataset of videos. You could also use existing libraries, such as OpenCV or scikit-learn, to help with the feature extraction and comparison steps.

OPS