Motion detection on series of static images needed

[login to view URL] is looking for a programmer / consultant to help develop a motion detection component within a video security system. about us: [login to view URL] offers an online video recording, monitoring and analysis services. We work with regular IP cameras, each 2 seconds a snapshot (.jpg file) is made and stored in Amazon S3. Then these snapshots must be processed with various algorithms to detect motion. We need a service that provides an XML Web Service (SOAP) API with a method which Receives a set of records where each record is: Id int - Snapshot Id CameraId int - Camera Id Url string - Snapshot S3 Url MdAlgorithm string - an identifier of the algorithm used MdArea int array[4] - a rectangle area on the image, top left, bottom right MdSettings - string, xml - settings specific to the algorithm used. The service downloads snapshots based on their Url, processes them with the algorithm and settings defined. Returns the result. The result is: A set of records where each record is: Id int - Snapshot Id MotionLevel number - a level of motion detected We expect from you to get a description of the algorithm you will use, its settings, what they mean, their allowed range of values. These settings will be passed in the MdSettings parameter as xml (xml structure will be negotiated) We also expect to hear from you the terms and the price Its much preferable that programming is done in .NET (C#). A typical snapshot dimensions is 640x480 with size of ~50Kb, however they may vary. From some cameras we receive snapshots 1024x768 with 180Kb size Motion detection is done on the series of images. We do not plan run MD on video yet. There is no limits on a type of movement. There can be anything - vehicles, people, animals, doors opening. As a result we should receive the fully compilable source code Please see the link below for more details [login to view URL]