Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

LOW-BANDWIDTH, LOW-COST VIDEO SURVEILLANCE

IP.com Disclosure Number: IPCOM000249955D
Publication Date: 2017-May-08
Document File: 4 page(s) / 122K

Publishing Venue

The IP.com Prior Art Database

Abstract

A system and method are proposed to reliably detect or monitor events of interests in an environment at low cost. The system consists of hundreds or thousands of small, low-cost, low resolution, solar-powered cameras with cellular connectivity and a cloud-based control center. Accordingly, each camera takes a few frames per minute and computes a semantic hash. If the hash is meaningfully different from the most recently reported, or is otherwise considered interesting, it sends the hash through a wireless connection to be uploaded in the cloud. The control center may analyze the hashes coming from all the cameras and may infer the events happening on the premises. The system may be deployed in areas far from civilization and may monitor environments such as traffic congestion, weather or for early detection of fire or trespassers of protected areas such as natural reserves or gas pipelines.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 47% of the total text.

LOW-BANDWIDTH, LOW-COST VIDEO SURVEILLANCE

ABSTRACT

A system and method are proposed to reliably detect or monitor events of interests in an

environment at low cost. The system consists of hundreds or thousands of small, low-cost, low

resolution, solar-powered cameras with cellular connectivity and a cloud-based control center.

Accordingly, each camera takes a few frames per minute and computes a semantic hash. If the

hash is meaningfully different from the most recently reported, or is otherwise considered

interesting, it sends the hash through a wireless connection to be uploaded in the cloud. The

control center may analyze the hashes coming from all the cameras and may infer the events

happening on the premises. The system may be deployed in areas far from civilization and may

monitor environments such as traffic congestion, weather or for early detection of fire or

trespassers of protected areas such as natural reserves or gas pipelines.

BACKGROUND

Video surveillance is a very powerful security technique. This allows videos to be stored

locally on the device and may be used to analyze any events of recent or past time. Online video

surveillance cameras stream the video through the internet to a cloud-based control center. This

approach is prohibitively expensive to set up at scale as it requires a wired or wifi connection to

the internet and a wired connection for electric supply. Further, the number of simultaneous

cameras which may be online is limited. Also, the quality of images from each camera may be

acceptable only for video surveillance inside a building or any location that is close by. For

monitoring vast or long areas like forests, savannas or gas pipelines, video surveillance

techniques are impractical. A competing approach may be to monitor any environment from the

sky using drones or satellites. They work the best at daytime and sunny weather. Performance of

drones and satellites is reduced considerably even in moderate mist, during night time or on

things covered by trees, tall grass, mountains or hills. Hence a cost-effective monitoring system

to detect events of interest in any environment is desirable.

DESCRIPTION

A system and method of multiple cameras that may operate to reliably detect or

monitor events of interest in an environment at low cost are proposed. The system as shown in

FIG. 1 consists of a number of small, solar-powered cameras with cellular connectivity

connected to a cloud-based control center such as a server. The cameras may be low-cost, low

resolution devices, to minimize the cost of the system, provided with cellular connectivity as

well as, in some cases, short range transmission such as WiFi or Bluetooth. Each camera

operates within a power budget to stay alive throughout the day and night. Accordingly, each

camera takes a few frames per minute and computes a semantic hash. If the hash is

meaningfully different from the most recently reported one or is otherwise considered

interesting, it sends the...