img_objects_detection use libtensorflow C API to detect objects from v4l2 usb camera, and use cairo library to draw object target