01 Riverbank Litter Accumulation

Scenario Description: River debris is often carried by currents and deposited on riverbanks. This system provides 24-hour real-time identification of litter accumulation along riverbanks, aiming to improve the efficiency of river cleanup and reduce the workload of environmental workers.

Applicable Scenarios: Applicable for identifying litter accumulation around riverbanks during daytime or nighttime with good lighting conditions.

Model + Post-processing: River Litter Detection

Camera:

Resolution: 1K and above, 4K and below

Target Image: Litter

Minimum Target Pixel Size: 75*75 pixels or more per 1K resolution

Lighting Conditions:

Illuminance: 90lux ≤ Illuminance ≤ 1200lux

Lighting Type: Color visible light

Minimum Pixel Monitoring Distance Limit (meters):

2.8mm

4mm

6mm

8mm

13m

20m

27m

Algorithm Effect Diagram:

02 River Floating Object Identification

Scenario Description: Identifies whether there are floating objects (domestic waste, water hyacinths, etc.) on the water surface in river scenarios. The algorithm provides 24-hour identification, aiming to improve the timeliness of river litter management and reduce the workload of environmental workers.

Applicable Scenarios: Applicable for identifying river floating objects during daytime or nighttime with good lighting conditions.

Model + Post-processing: River Floating Object Detection

Camera:

Resolution: 1K and above, 4K and below

Target Image: River Floating Objects

Minimum Target Pixel Size: 75*75 pixels or more per 1K resolution

Lighting Conditions:

Illuminance: 90lux ≤ Illuminance ≤ 1200lux

Lighting Type: Color visible light

Minimum Pixel Monitoring Distance Limit (meters):

2.8mm

4mm

6mm

8mm

13m

20m

27m

Algorithm Effect Diagram:

03 Area Intrusion

Scenario Description: Traditional perimeter defense often uses infrared beams, which have a high false alarm rate. Therefore, an area intrusion algorithm is introduced to eliminate interference from factors like animals and leaves, identifying illegal intrusion by personnel into electronic fences or restricted areas. This algorithm can also be used to prevent people from approaching critical areas.

Applicable Scenarios: Applicable for reservoir perimeter defense and critical area protection, suitable for surveillance during daytime and nighttime with supplementary lighting; typically used in scenarios where certain areas must be restricted from entry/intrusion/work by personnel.

Model + Post-processing: Pedestrian Detection + ROI Area Delineation

Camera:

Resolution: 1K and above, 4K and below

Target Image: Human figures

Minimum Target Pixel Size: 75*75 pixels or more per 1K resolution

Lighting Conditions:

Illuminance: 90lux ≤ Illuminance ≤ 1200lux

Lighting Type: Color visible light

Minimum Pixel Monitoring Distance Limit (meters):

2.8mm

4mm

6mm

8mm

13m

20m

27m

Algorithm Effect Diagram:

04 Fishing Activity Recognition

Scenario Description: By detecting the intersection of a person and a fishing rod, and prolonged inactivity, it determines whether someone is illegally fishing.

Applicable Scenarios: Applicable for detecting whether people fishing on the bank are doing so illegally during daytime.

Model + Post-processing: Pedestrian Detection, Fishing Rod Detection + ROI

Camera:

Resolution: 1K and above, 4K and below

Target Image: Personnel, Fishing Rods

Minimum Target Pixel Size: 75*75 pixels or more per 1K resolution

Lighting Conditions:

Illuminance: 90lux ≤ Illuminance ≤ 1200lux

Lighting Type: Color visible light

Minimum Pixel Monitoring Distance Limit (meters):

2.8mm

4mm

6mm

8mm

13m

20m

27m

Algorithm Effect Diagram:

05 Drone-based Swimming Detection

Scenario Description: Swimming is strictly prohibited in reservoirs due to deep water hazards. Drone-based visual swimming detection is introduced, triggering an alarm if someone is identified swimming.

Applicable Scenarios: Applicable outdoors with good lighting conditions, suitable for scenarios where swimming is strictly prohibited or for searching for people drifting in open water.

Model + Post-processing: River Detection + Swimming Detection

Camera:

Resolution: 1K and above, 4K and below

Target Image: Swimmers

Minimum Target Pixel Size: 75*75 pixels or more per 1K resolution

Lighting Conditions:

Illuminance: 90lux ≤ Illuminance ≤ 1200lux

Lighting Type: Color visible light

Minimum Pixel Monitoring Distance Limit (meters):

2.8mm

4mm

6mm

8mm

13m

20m

27m