Documentation for Axxon One 3.0/3.1 version. Documentation for Axxon One 2.0 version is also available.


On the page:


Detector configuration

To configure the License plate recognition XR detector, do the following:

  1. Go to the Detectors tab.
  2. Below the required camera, click Create…  Category: Vehicle  License plate recognition XR.

By default, the detector is enabled and set to recognize vehicle license plates.

If necessary, you can change the detector parameters. The list of parameters is given in the table:

ParameterValueDescription
Object features
Video stream

Main streamIf the camera supports multistreaming, select the stream for which detection is needed. Selecting a low-quality video stream reduces the load on the server.
Check in listsYesThe Check in lists parameter is disabled by default. If you want to use this detector to check in lists of vehicle license plates, select the Yes value (see Checking in lists of LPs)
No
Object identification
NameLicense plate recognition XREnter the detector name or leave the default name
EnableYesThe detector is enabled by default. To disable the detector, select the No value
No
TypeLicense plate recognition XRName of the detector type (non-editable field)
Decoder modeAutoSelect a processing resource for decoding video streams. When you select a GPU, a stand-alone graphics card takes priority (when decoding with NVIDIA NVDEC chips). If there is no appropriate GPU, the decoding will use the Intel Quick Sync Video technology. Otherwise, CPU resources will be used for decoding
CPU
GPU
HuaweiNPU
Basic settings
Algorithm for detecting vehicle directionBy LP coordinatesThe By LP coordinates algorithm is used by default. It works as follows: if the coordinates of the vehicle license plate shift from top to bottom (the vehicle moves towards the camera) and if the coordinates of the vehicle license plate shift from bottom to top (the vehicle moves away from the camera). Select the By vehicle scale change parameter if in the Surveillance window, the scale of the vehicle license plate increases (the vehicle moves towards the camera) and if the scale of the vehicle license plate decreases (the vehicle moves away from the camera)
By vehicle scale change
Frame size change1920

The analyzed frames are scaled down to a specified resolution (1920 pixels on the longer side) by default. The value must be in the range [640, 10000]. The following algorithm is used:

  • If the longer side of the source image exceeds the value specified in the Frame size change field, it is divided by two.
  • If the resulting resolution falls below the specified value, the algorithm stops, and this resolution will be used further.
  • If the resulting resolution still exceeds the specified value, it is divided by two until it is less than the specified resolution.

Note

For example, the source image resolution is 2048*1536, and the specified value is set to 1000. In this case, the source resolution will be halved two times (512*384), as after the first division, the number of pixels on the longer side exceeds the specified value (1024 > 1000). If detection is performed on a higher resolution stream and detection errors occur, we recommend reducing the compression.

Initiate event in (sec)0

Specify the time in seconds after which an event must be generated after recognizing a license plate in the frame. The value must be in the range [0, 60]


Effect of values:

  • 0an event is generated when the vehicle leaves the detection area. The best result is selected from the entire observation period.
  • 1 or greater—an event is forced to be generated N seconds after the license plate is first detected, even if the vehicle is still in the frame.

Recommendations:

  • 0—the best detection is at the END of the area (vehicle is moving away, rear license plate is visible).
  • 12—the best detection at the BEGINNING of the area (vehicle is approaching, front license plate is visible).
  • Mean value— bidirectional traffic in one lane.

Impact on resources:

  • No 
LP recognition basic settings
Countries(Select all), Afghanistan, Albania, Algeria, American Samoa, Andorra, Angola, Anguilla, Antigua and Barbuda, Argentina, Armenia, Aruba, Australia, Austria, Azerbaijan, Bahamas, Bahrain, Bangladesh, Barbados, Belarus, Belgium, Belize, Benin, Bermuda, Bhutan, Bolivia, Bosnia, Botswana, Brazil, Brunei, Bulgaria, Burkina Faso, Burundi, Cambodia, Cameroon, Canada, Cape Verde, Cayman Islands, Central African Republic, Chad, Chile, Chine, Christmas Island, Cocos Isands, Colombia, Comoros, Congo, Cook Islands, Costa Rica, Croatia, Cuba, Cyprus, Czech Republic, Denmark, Djibouti, Dominica, Dominican Republic, East Timor, Ecuador, Egypt, El Salvador, Equatorial Guinea, Eritrea, Estonia, Ethiopia, Falkland Islands, Faroe Islands, Fiji, Finland, France, French Guiana, French Polynesia, French Southern Territories, Gabon, Gambia, Georgia, Germany, Ghana, Gibraltar, Great Britain, Greece, Greenland, Grenada, Guadeloupe, Guatemala, Guinea, Guinea Bissau, Guyana, Haiti, Holy See, Honduras, Hong Kong, Hungary, Iceland, India, Indonesia, Iran, Iraq, Ireland, Israel, Italy, Ivory Coast, Jamaica, Japan, Jordan, Kazakhstan, Kenya, Kiribati, Kosovo, Kuwait, Kyrgyzstan, Laos, Latvia, Lebanon, Lesotho, Liberia, Libya, Liechtenstein, Lithuania, Luxembourg, Macau, Macedonia, Madagascar, Malawi, Malaysia, Maldives, Mali, Malta, Marshall Islands, Martinique, Mauritania, Mauritius, Mayotte, Mexico, Micronesia, Moldova, Monaco, Mongolia, Montenegro, Monserrat, Morocco, Mozambique, Myanmar, Namibia, Nauru, Netherlands, Netherlands Antilles, New Caledonia, New Zealand, Nicaragua, Niger, Nigeria, North Korea, Northern Mariana Islands, Norway, Oman, Pakistan, Palau, Palestine, Panama, Papua New Guinea, Paraguay, Peru, Philippines, Pitcairn Island, Poland, Portugal, Puerto Rico, Qatar, Reunion Island, Romania, Russia, Rwanda, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Samoa, San Marino, Sao Tome and Principe, Saudi Arabia, Senegal, Serbia, Seychelles, Sierra Leone, Singapore, Slovakia, Slovenia, Solomon Islands, Somalia, South Africa, South Korea, South Sudan, Suriname, Spain, Sri Lanka, Swaziland, Sweden, Switzerland, Syria, Taiwan, Tajikistan, Tanzania, Thailand, Tibet, Timor Leste, Togo, Tokelau, Tonga, Trinidad and Tobago, Tunisia, Turkey, Turkmenistan, Turks and Caicos Islands, Tuvalu, Uganda, Ukraine, United Arab Emirates, United Kingdom, United States, Uruguay, Uzbekistan, Vanuatu, Vatican, Venezuela, Vietnam, Virgin Islands British, Virgin Islands USA, Wallis and Futuna Islands, Western Sahara, Yemen, Zambia, Zimbabwe

The Not used value is selected by default. In the window that opens, set one or several checkboxes next to the countries whose license plates must be recognized. Click the Apply button. To cancel the selection of countries, click the Cancel button

Country name 1 (deprecated)(Not used), Afghanistan, Albania, Algeria, American Samoa, Andorra, Angola, Anguilla, Antarctica, Antigua and Barbuda, Argentina, Armenia, Aruba, Australia, Austria, Azerbaijan, Bahamas, Bahrain, Bangladesh, Barbados, Belarus, Belgium, Belize, Benin, Bermuda, Bhutan, Bolivia, Bosnia, Botswana, Brazil, Brunei, Bulgaria, Burkina Faso, Burundi, Cambodia, Cameroon, Canada, Cape Verde, Cayman Islands, Central African Republic, Chad, Chile, Chine, Christmas Island, Cocos Isands, Colombia, Comoros, Congo, Cook Islands, Costa Rica, Croatia, Cuba, Cyprus, Czech Republic, Denmark, Djibouti, Dominica, Dominican Republic, East Timor, Ecuador, Egypt, El Salvador, Equatorial Guinea, Eritrea, Estonia, Ethiopia, Falkland Islands, Faroe Islands, Fiji, Finland, France, French Guiana, French Polynesia, French Southern Territories, Gabon, Gambia, Georgia, Germany, Ghana, Gibraltar, Great Britain, Greece, Greenland, Grenada, Guadeloupe, Guam, Guatemala, Guinea, Guinea Bissau, Guyana, Haiti, Holy See, Honduras, Hong Kong, Hungary, Iceland, India, Indonesia, Iran, Iraq, Ireland, Israel, Italy, Ivory Coast, Jamaica, Japan, Jordan, Kazakhstan, Kenya, Kiribati, Kosovo, Kuwait, Kyrgyzstan, Laos, Latvia, Lebanon, Lesotho, Liberia, Libya, Liechtenstein, Lithuania, Luxembourg, Macau, Macedonia, Madagascar, Malawi, Malaysia, Maldives, Mali, Malta, Marshall Islands, Martinique, Mauritania, Mauritius, Mayotte, Mexico, Micronesia, Moldova, Monaco, Mongolia, Montenegro, Monserrat, Morocco, Mozambique, Myanmar, Namibia, Nauru, Netherlands, Netherlands Antilles, New Caledonia, New Zealand, Nicaragua, Niger, Nigeria, Niue, North Korea, Northern Mariana Islands, Norway, Oman, Pakistan, Palau, Palestine, Panama, Papua New Guinea, Paraguay, Peru, Philippines, Pitcairn Island, Poland, Portugal, Puerto Rico, Qatar, Reunion Island, Romania, Russia, Rwanda, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Samoa, San Marino, Sao Tome and Principe, Saudi Arabia, Senegal, Serbia, Seychelles, Sierra Leone, Singapore, Slovakia, Slovenia, Solomon Islands, Somalia, South Africa, South Korea, South Sudan, Suriname, Spain, Sri Lanka, Sudan, Swaziland, Sweden, Switzerland, Syria, Taiwan, Tajikistan, Tanzania, Thailand, Tibet, Timor Leste, Togo, Tokelau, Tonga, Trinidad and Tobago, Tunisia, Turkey, Turkmenistan, Turks and Caicos Islands, Tuvalu, Uganda, Ukraine, United Arab Emirates, United Kingdom, United States, Uruguay, Uzbekistan, Vanuatu, Vatican, Venezuela, Vietnam, Virgin Islands British, Virgin Islands USA, Wallis and Futuna Islands, Western Sahara, Yemen, Zambia, Zimbabwe



Attention!

Starting with Detector Pack 3.17.1, this parameter is deprecated. Its value is only taken into account if you select the Not used value in the Countries parameter. Otherwise, this parameter is ignored.

Albania is selected by default. From the drop-down list, select one country whose license plates are actually found in the detection area and which need to be recognized

 

 

 

 

Country name 2 (deprecated)
Country name 3 (deprecated)
Country name 4 (deprecated)
Country name 5 (deprecated)
LP recognition advanced settings
Period of ignoring repeated recognitions (sec)0

Specify the period of time in seconds during which repeatedly recognized license plates are ignored. The value must be in the range [0, 86400].


Effect of values:

  • 0 (the default value)—history is disabled. Each recognition generates a separate event.
  • 159 seconds—short-term filtering. One license plate can generate no more than one event during the specified interval.
  • 60300 seconds—medium filtering. Optimal for checkpoints and areas where vehicles can be stopped.
  • 3600 seconds and greater—long-term filtering (hours). Suitable for entrance logs where only the first vehicle appearance is recorded.
  • 86400 seconds (maximum)—one license plate, one event per day.

Recommendations

  • 0 seconds—parking lots with short stops, each entry and exit must be recorded.
  • 60300 seconds—checkpoints, elimination of duplicates of parked vehicles.
  • 3600+ seconds—access control to the area, only the first entry is important.

Impact on resources:

  • Minimum memory usage for storing the history of recognized license plates 
Maximum Levenshtein distance1

Specify the maximum Levenshtein distance between license plate characters. The value must be in the range [0, 999].

This parameter is used to determine what makes a license plate different from the previous license plate (Levenshtein distance). For two identical readings, the Levenshtein distance is zero. This parameter helps to reduce false positives when a license plate is read several times before a result is given and a partial reading still represents a valid license plate.

Example: real license plate is "AB-4567".

  • reading 1: AB-4567
  • reading 2: AB-456

If the maximum distance is set to 0 (exact match), the system reports two license plates. But if the distance is set to 1 (or greater), the system reports the first license plate, but not the second one. This is because a license plate with one distinct character is considered a readout of the same license plate, that is, a probable misreading has occurred

Note

  • For the countries where the minimum number of characters in license plates is greater than 5, we recommend setting the value to 2 or greater.
  • For the countries where license plates can have 5 or fewer characters, set this value to 1.
Minimum height of license plate characters (pixels)14Specify the minimum height of license plate characters in pixels. This is the distance between the top pixel of the character and the bottom pixel. The value must be in the range [14, 70]
Maximum height of license plate characters (pixels)70

Specify the maximum height of license plate characters in pixels. The value must be equal to or greater than the value in the Minimum height of license plate characters (pixels) parameter. The value must be in the range [34, 70]

Note

  • For license plates with different character sizes, such as those of Middle Eastern countries, the maximum height is applied to the largest set of characters found in the license plate.
  • The optimal character size for recognition is 22–24 pixels.
  • Characters less than 18 pixels high increase the likelihood of misrecognition, as details of the character shape can be lost, especially if the image quality is poor: a 'G' can turn into a 'C', and so on.
  • Characters more than 55 pixels high can take much longer to process.
Minimum authenticity threshold80

Specify the minimum value of recognition authenticity as a percentage. The value must be in the range [1, 100]. The parameter determines the minimum average confidence level of all characters of a license plate. 

If the average value of all characters has a confidence level below the set threshold, the entire license plate is rejected and no event is generated


Effect of values:

  • 5070% (low threshold)—more license plates are accepted in general. License plates with individual poorly readable characters can pass filtering.
  • 7085% (medium threshold)—balanced filtering. Recommended for most use cases.
  • 8595% (high threshold)—only license plates with a high degree of confidence are accepted. Recommended for high-quality images and countries with open grammars (for example, the USA).

Difference from the Minimum character confidence threshold parameter

  • Minimum authenticity threshold (overall threshold) checks the average confidence value of all characters. Example: character confidence values ​​are [95, 95, 60, 95, 95, 95]. The average value is 89%. The license plate is accepted if the threshold is set below 89%.
  • Minimum character confidence threshold (character threshold) checks each character individually. The same license plate is rejected if the threshold is set above 60%, as one character is below the threshold.

Recommendations:

  • 5070%—difficult shooting conditions.
  • 7085%—standard conditions.
  • 8595%—high-quality images, USA (custom license plates).

Impact on resources:

  • No
Minimum character confidence threshold80

Specify the minimum threshold value of license plate recognition confidence as a percentage. The value must be in the range [1, 100]. The parameter determines the minimum level of recognition confidence for each individual character of the license plate.

If at least one character has a confidence level below the set threshold, the entire license plate is rejected and no event is generated.


Effect of values:

  • 5070% (low threshold)—more license plates are accepted, including those that are difficult to read. The risk of character recognition errors increases (for example, O ↔ 0, B ↔ 8, and so on).
  • 7085% (medium threshold)—balanced filtering. Recommended for most use cases.
  • 8595% (high threshold)—only characters with a high degree of confidence are accepted. The number of errors decreases, but the risk of skipping dirty or damaged license plates increases.

Recommendations

  • 5070%—difficult conditions: dirt, damage, poor lighting.
  • 7085%—standard conditions.
  • 8595%—high-quality images, stable lighting.

Impact on resources:

  • No
Country pattern compliance
YesThe parameter is enabled by default. To disable the check of country pattern compliance, select the No value
No
Extra processing for LP searchYes

The parameter is disabled by default. To enable the extra processing for license plate search, select the Yes value


Effect of values:

  • No—the standard algorithms of license plate recognition are used. Reduces the CPU load.
  • Yes—the advanced detection algorithms are used. Improves the detection of non-standard, damaged, or dirty license plates. Increases the CPU load.

Recommendations:

  • Enable (the Yes value)—for dirty license plates, non-standard sizes, complex angles, low image contrast.
  • Disable (the No value)—high CPU load, standard conditions.

Impact on resources:

  • Increased CPU load when this option is enabled
No
Minimum number of characters in LP2Specify the minimum number of characters in a license plate. The value must be in the range [2, 16]
Maximum LP slope angle40Specify the maximum slope angle of the license plate. The value must be in the range [0, 999]
Recognition algorithm complexity


Low level

Select the complexity level of recognition algorithm from the list. The Medium level is selected by default

Note

  • The higher the complexity of the algorithm, the longer the processing time.
  • Depending on the images, the difference between the low and high complexity levels can be up to 4%, but on average the difference is about 2 or 3%.
  • If the image is of high quality and license plates have a standard shape, the difference between medium and high level doesn't matter. You can set the Medium level to make license plate processing faster.
  • We recommend using the Low level only in complex scenes, low computing power, and when vehicle detection is preferred over license plate recognition.
Medium level
High level
Minimum number of same LP readings1

Specify the minimum number of readings of the same license plate.

For example, if you set the value to 2, you will receive a massage about a license plate only if two identical readings have been made. The value must be in the range [1, 999].


Effect of values:

  • 1 (the default value)—the license plate is returned after the first reading. Provides maximum coverage but increases the likelihood of false events.
  • 2—the license plate must be read at least twice. Significantly reduces the number of false events. The license plate can be skipped if the vehicle quickly leaves the field of view.
  • 3 or greater—high filtering level. Increases the risk of skipping license plates in fast-moving traffic.

Recommendations

  • 1—fast traffic, skipping license plates is forbidden.
  • 2standard conditions, optimal balance of accuracy and completeness. Particularly recommended for the USA (custom license plates).
  • 3 or greater—only for slow traffic, when accuracy is critical.

Impact on resources:

  • No 
Maximum number of same LP readings10

Specify the maximum number of readings of the same license plate. The value must be in the range [1, 999]. When the set limit is reached, the result is returned immediately, without waiting for the timeout to expire.


Effect of values:

  • Low (15): Results are returned faster. Fewer reads are performed to select the best option, potentially resulting in lower accuracy.
  • Medium (515): Optimal balance between speed and accuracy.
    Recommended for most use cases.
  • High (15 or greater): More reads are performed to select the best result. Result latency increases.

Recommendations:

  • 510—standard moving traffic.
  • 1020 — slow traffic, barriers, checkpoints.

Impact on resources:

  • The higher the value of the parameter, the higher the CPU load. 
Recognition timeout (sec)0

Specify the recognition timeout in seconds. The value must be in the range [0, 999]. The parameter determines the period during which multiple readings of the same license plate are accumulated before the final decision is made. After the timeout, the system selects the best result from all accumulated readings.


Effect of values:

  • 0 seconds—the result is generated immediately after the first reading. Minimum memory usage, minimum accuracy.
  • 12 seconds—short accumulation period. Suitable for fast traffic (100 km/h and above).
  • 35 seconds—standard accumulation period. Optimal balance of speed and accuracy for most use cases.
  • 5 seconds or greater—long accumulation period for maximum accuracy. Recommended for slow-moving traffic, areas with barriers, and checkpoints. Increases memory usage.

Interaction with the Maximum/Minimum number of same LP readings parameters

  • If the timeout expires but the minimum number of readings has not been reached, the result is discarded.
  • If the maximum number of readings is reached before the timeout expires, the result is returned immediately.

Recommendations:

  • 0 seconds—only for static scenes or external triggers.
  • 12 seconds—highways, fast traffic.
  • 35 seconds—standard city traffic.
  • 5 seconds or greater— slow traffic, barriers, checkpoints.

Impact on resources:

  • The higher the value of the parameter, the higher the memory usage
Vehicle recognition basic settings
Vehicle recognitionYesThe parameter is disabled by default. To enable the vehicle recognition, select the Yes value
No
Vehicle recognition advanced settings
Minimum vehicle recognition confidence threshold50Specify the threshold value of the minimum vehicle recognition confidence as a percentage. The value must be in the range [1, 100]
Vehicle analytics qualityMedium levelSelect the quality of the vehicle analysis from the list. The Medium level is selected by default
High level
Maximum level

By default, the entire frame is analyzed. If you want to analyze only a part of a frame, in the surveillance window, specify the area in which detection and recognition of vehicles will be performed. You can adjust the area by moving the anchor points

Note

  • For convenience of configuration, you can "freeze" the frame. Click the button. To cancel the action, click this button again.
  • To hide the detection area, click the button. To cancel the action, click this button again.
  • To delete the selected area, click the button.

To save the parameters of the detector, click the Apply  button. To cancel the changes, click the Cancel  button.

Configuration of the License plate recognition XR is complete.

Examples of detector configuration for solving typical tasks

  • Fast traffic (highways, 80 km/h or more):
    ParameterValue
    Initiate event in (sec)0
    Minimum number of same LP readings1
    Maximum number of same LP readings5
    Recognition timeout (sec)12
  • Standard city traffic:
    ParameterValue
    Initiate event in (sec)0
    Minimum number of same LP readings2
    Maximum number of same LP readings10
    Recognition timeout (sec)35
  • Checkpoints, barriers (slow traffic):
    ParameterValue
    Initiate event in (sec)12
    Minimum number of same LP readings23
    Maximum number of same LP readings1520
    Recognition timeout (sec)5 or greater
    Period of ignoring repeated recognitions (sec)60300
  • Parking lot:
    ParameterValue
    Initiate event in (sec)0
    Minimum number of same LP readings2
    Maximum number of same LP readings10
    Recognition timeout (sec)35
  • Difficult conditions (dirt, damage, poor lighting):
    ParameterValue
    Extra processing for LP searchYes
    Minimum character confidence threshold5070
    Minimum authenticity threshold5070
  • No labels