ALL RIGHTS FOR THE PHOTOES RESERVED TO Edsel L's photostream, taken from flickr ! keep the credit

Our ideas, concepts, and projects on how RDV Tech applies computer vision solution.

We are working mainly in these three areas below, but not limited to an expansion in other areas of implementations in the near future.

security

Auto-alert video surveillance

Unlike the common CCTVs for video surveillance, our system create an alarm to automatically detect objects that pose a security threat, based on client’s security risks. After we define region of interest (ROI) on CCTV camera to set up the object detection area, then we configure the alarm rules. That alarm may detect specific objects according to their characteristics, such as size, movement, color, and speed. If the specified object is detected then the system will alert the user by an automated alarm in real-time. <Concept>

Automatic number plate recognition

Our number plate recognition system is built for an automated car-parking system. Although there are many commercially available automatic number plate recognition now, but RDV Tech has the ability to create one that fits specific requirements for car-number plate in Indonesia and suit the parking system in Indonesian offices or malls. With this automatic number plate recognition system embed in car-parking system, we can reduce the use of parking-ticket-papers and the risk of parking ticket lost. <Concept>

lifestyle

Integrated fruit and veggie recognizer

At most supermarkets, fruits and vegetables are weighed at the weighing kiosk by supermarket assistants or at the cashier. Sometimes it may cause a long queue for weighing or you might end-up waste your time to wait for the supermarket assistant to come and weight.

With this integrated fruit and veggie recognizer scale, you can measure your groceries yourself. You just have to put the fruit or vegetable on the scale and no need to choose which type of fruit/veggie button. Wait for the built-in camera on the scale to scan your products. Then, it will recognize which fruit or vegetable you buy and give the sticker-price according to its type and weight. Moreover, we can offer you a recipe according to what fruit/veggie you bought. <Concept>

healthcare

Automatic segmentation of heart muscle post heart-attack

Accurate identification of the tissue or regions in the organ to be studied is the first step in medical image analysis. Using automatic segmentation, we identify and model tissues with sub-voxel accuracy. One of our projects is the automatic segmentation of three different tissues in myocardical infarction of myocardium (heart muscle) after heart attack.

Using statistics and machine learning methods, we can detect and quantify those three regions from MRI images. The quantification in area and volume are then used for assessing the severity of heart-attack and determinung the most suitable therapy for the patients. This software module is intended to be used by cardiologists and nurses in hospitals. <Project>