What are the system requirements and how to use Analytic Search.
The primary requirement is that the camera supports ONVIF M and is equipped for People/Vehicle object classification. This functionality is controlled via a feature flag within the Qumulex Control software and requires the appropriate licensing to activate. For details on compatibility and pricing, contact Sales@qumulex.com.
The initial step in this process is to enable the capability by activating the QXP-A License. Navigate to the Camera Tab, select Edit Camera, and scroll to the bottom of the page. Toggle on Enable Profile M Metadata, then ensure you click Save Camera to apply the changes.
Step 1 is scope to the highest place in the hierarchy for your Analytic search, this can be narrowed from the Analytics menu later if applicable.
Step 2 is from the Search Tab, select the Portrait icon at the top of the page. This is the new Analytics Page.
On the Analytics page, there are several time-related options. You can begin a search from the current moment by clicking the Now button at the top of the page. Alternatively, use the Calendar function on the left side to select a specific date range. Additionally, you can adjust the search scope here to focus on a particular site or location.
In the upper right corner, you'll find the Scaling slider, which adjusts the display based on your screen size. Use the sliding scale to make snapshots on the page appear smaller or larger, depending on your preference.
To initiate an Analytic Search, click Add Analytic and select a type—Face, Human, or Vehicle.
Depending on the make and model of the device, the same attributes may be classified under both Face and Human. To ensure comprehensive results, you can create a search for either category.
In this example I am using the Human category, and I am searching for an individual with a Red - Top Color.
I would then click Search
From my initial search, the results include everything the camera classified as red or with a reddish tint. Several factors influence the camera's classification or may result in false positives. These factors include lighting conditions, the object's distance from the camera, and ensuring the device is running the latest and most optimized application from the manufacturer.
In this scenario, the objective is to locate an individual wearing a bright red shirt. To increase the likelihood of finding relevant results, the initial filter was set to include items with a likelihood range of 1%–99%. To narrow the search further, the filter can be adjusted to focus on results within the 90%–99% likelihood range for red shirt matches. Select the Likelihood category marker and adjust the slider to 90%. Once updated, click Search to generate the new filtered results.
With my new Search, I now see the individual in question.
The final step in the search process is selecting a thumbnail to review the recorded video. Clicking on a result will open the video player on the right-hand side of the screen, which will begin playback from the moment the individual was detected. From here, you can continue browsing through thumbnails or conclude your investigation by creating a video clip for export or sharing.
Operators can share search results with other Qumulex users by clicking the Share button at the top of the page. Copy the search parameter string to the clipboard and paste it into another browser's URL bar to replicate the previously performed search.