Abstract: Abstract: Small to medium-sized seeds are characterized by non-food crops with an average diameter of less than 7 mm, mainly including rape, wheat，alfalfa, panax ginseng, and cereals. Since crop cultivation is gradually changing from the original grain to the diversified economic type under the adjustment of agricultural industry in China, the small to medium-sized seed crop is generally expected to serve as a sunrise industry with more promising markets. However, it is difficult to monitor the sowing status because the seeder seeds thought tube is enclosed status, when small and medium-sized seeds are sown in the field. In addition, in the detection of small to medium-sized seed crops, the seeds sow frequency is higher compared with large seed crops (rape row frequency is about 20-40 Hz, wheat row frequency is about 100-300 Hz), and it is difficult to accurately distinguish single seeds because of the high row frequency and poor seed sequencing, and there are also difficulties in detecting small to medium-sized seeds because of the small seed size and the weak seed signal generated when the seeds cross the sensing area. The above reasons make it difficult to accurately detect small and medium sized seeds. Correspondingly, the quality of seeding depends seriously on the seed rower, where missing seeds can inevitably result in sparse seedlings and broken strips in the complex field environment. Once a missed sowing cannot be found in time, a large area to be missed can inevitably occur, especially for the wide high-speed seeder with a high operation speed and sowing width. Manual replenishment of seedlings at a later stage is also time-consuming and labor-intensive. Therefore, it is necessary to real-time monitor the sowing process of small and medium-sized seeds for the high quality of crop sowing. Alternatively, the state-of-art technology of seeding detection is widely expected to realize variable seeding in intelligent agriculture. In this study, a systematic review was performed on the progress of detection technology and corresponding equipment for small and medium-sized seeds in the international market. Five sensing methods were selected to clarify the benefits and limitations, including mechanical electromechanical alarm, machine vision, photoelectric, capacitive, and piezoelectric sensing detection. Meanwhile, seven evaluation indicators were selected to determine the sowing detection, including sower failure monitoring, sowing volume, sowing frequency, rowing spacing, planting spacing, missing sowing, reseeding. The specific indexes of sowing detection were proposed for different sowing modes in combination with precision agricultural requirements. Furthermore, a full overview was made on the leakage compensation and information transmission of sowing detection to avoid the sowing leakage. In addition, a systematic summary was given on the sowing detection related technologies, and the possible development requirements for small and medium-sized seed sowing detection in the context of precision agriculture. Future detecting trend was foreseen when sowing small and medium-sized seeds. Consequently, three aspects were elaborated, including the current detection, detection indexes, and sowing detection for small and medium-sized seeds. Technical difficulties were given using detection devices for small and medium-sized seed flow during this stage. The finding can provide a key technical support to sowing detection of small and medium-sized seeds.