This technology is mainly used in the monitoring system of image analysis, the input image data through the low complexity of Foreground Detection for front and back separation and background modeling. The resulting foreground information will send to Object Labeling module to assign individual label values to the foreground connectivity block, but when the current background image information is too close, the foreground data will be too broken. So use the Object Grouping to achieve the complete object. Object Tracking will match the object label of the previous frame Information, if it already exists then take replace, and that will achieve tracking results. However, this system must exist some occlusion case, So there are corresponding solutions to the various possible scenarios to achieve good tracking.
Real-time lossless compression not only improves diagnostic accuracy but also for subsequent medical applications. We use Adaptive Linear Prediction to reduce the dynamic prediction error, and use Content - Adaptive Golomb rice coding to enhance the overall system compression rate.

本技術是一種未來智慧控制系統的模擬應用，相較於市面上的IoT方案大多著點在智慧手機為媒介，我們主要著重在於行為本身的操作，可以避免攜帶式的不便，並且同時擁有更多的未來方案的可行性，在鏡頭模組在居家的重要性逐漸增加的現況中，行為控制與辨識將會是更符合智慧家庭需求的方案。

可用於偵測防盜、人形追蹤、寵物等，也能結合一些行為辨識的處理，能夠達到一些智慧控制系統的應用，或是結合物件軌跡與Direct Linear Transformation，即可達到室內定位的成果。