katt / vesper ✧

FLOCK patent indicates it’s a national citizen identification and tracking system using neural networks. The ALPR cameras are just an entry point. The real product is the DSOQ database and the AI models that identify, classify, and track people. 🚨👀 The "opt-in" language and the dynamic footprint language prove this was never just about police cameras. This was designed from day one to incorporate private cameras, smartphones, and any video source. Its a surveillance architecture designed to crowdsource video from the public.🚨👀 —- And if you look at the NEURAL NETWORK Figure you see The patent explicitly includes a neural network for identifying PEOPLE It explicitly includes a neural network for HEIGHT AND WEIGHT. Those are body characteristics. Those are biometric surrogates. You do not need someone's height, weight, and clothing type to read a license plate. You need those things to identify a PERSON when you cannot or choose not to use facial recognition. —- Not only this most disturbing is🚨👀 The patent defines "fingerprint data" as including "Hough lines, non-zero pixel ratio, black/white pixel ratio, white balance, hash value, neighborhood data, and planar, induced subgraph of all relevant objects. " Read that list carefully🚨👀. Hough lines measure geometric shapes. Non-zero pixel ratios measure object density. Black/white pixel ratios measure contrast patterns. White balance measures color temperature. Hash values create unique digital identifiers. Planar subgraphs map relationships between objects. human body metric data. Combined with the height/weight neural network and the clothing type neural network, the fingerprint data creates a unique biometric profile of a person that can identify them WITHOUT their face being visible. You can identify someone by their height, build, clothing, movement pattern, and neighborhood location without ever seeing their face.🚨👀 —- Even more disturbing 🚨😬 The patent explicitly states: "embodiments of the system are not constrained by any geographic footprint associated with a set network of surveillance devices" and can "leverage a dynamic geographic footprint defined by the geographic locations of unrelated surveillance devices." Translation: the system was designed to expand its coverage by incorporating ANY camera, ANYWHERE. The patent envisions neighborhood watches, commercial security systems, home security systems, traffic cameras, AND portable recording devices which means smartphones, body cameras, drones, and any personal video device. The "dynamic" footprint is not a feature. It is a design specification for national-scale coverage using every available camera as a node.🚨👀⚠️⚠️⚠️ —- 🚨⚠️👀 Read FIG. 7. The historical analysis module "compares outputs, decoded symbology and/or confidence ratings of seed record with other records in the DSOQ database" and "returns records statistically significantly similar to the seed record." It has tracking function. Given one sighting of a person or vehicle at one location and time, the system queries the entire database for ALL other records that are statistically similar. It finds every other time that person or vehicle appeared on ANY camera in the network. It reconstructs movements across the entire grid. The patent calls this "historical analysis." The correct term is movement reconstruction.🚨⚠️👀 @RepThomasMassie patentimages.storage.googlea… patents.google.com/patent/US…