To realize the continuous compaction control (CCC) of asphalt pavement during construction, continuous detection method was investigated for the compaction degree values. For the trajectory of rollers, a collaborative positioning method was proposed. For the monitoring of rolling process, an embedded-based detection system was designed. For the evaluation of rolling effect, harmonic analysis was introduced and a new index, vibration compaction energy value (VCVe), was proposed. Positioning experiments were conducted, and the accuracy was improved to 0.48 m. Rolling tests were performed, and typical compaction meter values (CMVs), compaction control values (CCVs), and VCVe were obtained. The referenced compaction degree by conventional way was 94.6%, which was used to calibrate the detected values of compaction degree indexes. The results showed that continuous compaction detection can be achieved based on positioning system and vibration analysis. Compared with CMV and CCV, VCVe is less discrete, more stable, and consistent to describe the compaction state. Though, all the CMV, CCV, and VCVe indexes are unable to be used for quality assurance directly or alone, they could be an aid for quality control. Continuous compaction detection system meets the monitoring requirements of pavement construction at a lower cost and could lay a foundation for the intelligent compaction (IC).