Object localization and position estimation is a demanding need in robotic, military and automation applications and fields. GPS is an effective solution for outdoor localization however due to radio interference it is less applicable in indoor environments and it also suffers from limited accuracy. Other RF positioning systems have also been developed to overcome these issues such as techniques based on UWB. In parallel with RF techniques, vision methods have received considerable attention for object tracking applications. The implementation of such algorithms on sensor-processor combinations tends to be power consuming and slow. As an alternative, smart vision sensors provide the opportunity for low power high speed image processing. In this thesis two modulated light detection smart CMOS image sensors are presented. The designs have the ability to sense asynchronous signals transmitted from electronic markers such as flashing light emitting diodes (LEDs) tagged on moving robots. With the presented sensors, object localization and position detection functions are simplified, performed at higher speeds in real time, and power requirement is reduced. The sensors in-pixel processing, filters out the background image data, detects the modulated marker regions and produces a single bit result for each pixel which represents the existence or absence of modulated signal in that pixel. Key Words Smart CMOS sensor, Modulated light, Electronic light tag, Vision, Localization.
