
Smart monitoring target camera
Services:
Project overview
The project is focused around providing in-depth research for a known company in the ammunition industry. Our client designs equipment for the shooting community, in particular target cameras significantly enhancing long range shooting events.
Abto Software originally proposed the thought of implementing computer vision for faultless hole recognition. But having carefully tested the application, our team has discovered that the currently utilized UI/UX approach is confusing even with manual markers and provides no room for adopting computer vision.
After conducting comprehensive analysis, we came up with “smart monitoring”, a concept centered around dispatching current design issues, rethinking existing, inefficient features, and leveraging advanced technology, in particular CV techniques.
Main goals
The project’s main objective was the meticulous investigation of the legacy solution’s strong and weak points. To achieve maximum insight and accuracy, our team precisely reproduced the application of the target camera at the shooting range.
Abto Software was mainly focused on:
- Preliminary and in-depth investigation
- Issue and strategy definition
- Concept devising
- POC composure
Main challenges
The project’s main challenge was the incomplete understanding of the basic needs and problems of end-users. To obtain that understanding, our team conducted thorough field testing.
At the very beginning of the discovery stage, we faced several challenges:
1. The setup of an applicable environment
Having started the research, we found that the legacy application can’t be utilized without special equipment. To resolve this problem, we ordered a set of four specialized cameras.
2. The need to go further beyond conventional approaches
When running the investigation, we realized that simulating practical application isn’t bringing desired results. To obtain deeper insight and accuracy, we went far beyond accustomed methods by allocating extra resources to conduct field testing outside the office environment.
This way, we identified several issues related to:
- Timestamp placing
- Hole marking
- Target and result visibility
- Shooting history
Having reproduced the conditions of utilization, we could better understand those issues in detail:
- The shooter must manually record events, which might cause loss of progress in case of distractions
- The shooter must then mark each new hole, which is not only quite confusing, but might cause errors
- The shooter might miss ground shots because of dust clouds when switching his attention to the manual input
- The shooter can’t review historical data, which complicates tracking progress
- Since the displayed image remains static, the user had difficulties with distinguishing real-time events from screenshots reviewing results
- Since there’s playback delay, the user isn’t able to track well-hit shots in accuracy
And multiple accompanying problems related to visual and audio factors, as well as the special needs of the individual user, for example, right- and left-hand use.
Our team set out to create a concept, which would completely bypass the shortcoming the user was facing. Having gathered enough information, we devised an interface, which helps clearly understand every event being displayed on screen.
Core functionality
At the very moment, the existing legacy application provides the following functionality:
- Live pictures from the target camera with a certain FPS
- Saved pictures (directly to the gallery)
- Video recording made directly from the device screen, thus limited by the device resolution
- Manual timestamps
- Frame switching to compare last snapshot with the real-time picture (blinker mode)
- Distance setting to calculate shooter-target distance
The application provides multiple essential features for both shorter and extreme range shooting events. Despite this, our team identified problems notably impacting user experience.
Our concept is aimed at providing smart monitoring along with additional functionality:
- By using computer vision along with sound detection, the solution can automatically make timestamps when recognizing relevant events:
– New holes (the app will constantly compare frames and highlight the differences)
– Ground shots (the app will display dust clouds directly on the screen)
– Loud sounds (the system might distinguish rifle sounds to filter other shooters)
- The app can assist the shooter in viewing the recording in a timeline manner:
– The user can reproduce relevant events (rifle sounds, dust clouds, and other)
– The user can slide the recording to find particular events
– The user can go to next and previous events
- The app can provide the opportunity to change historical data:
– Mark holes if the engine fails
– Mark holes in a different color if the system confuses rifle sounds
– Make additional text/voice notes - Right- and left-hand use
At the same time, the application shouldn’t recognize irrelevant events:
- Grass movements
- Target movement
- Bugs, birds, small animals
- And other potential distractions
Our contribution
After signing the mutual NDA agreement, our team:
1. Conducted a preliminary discovery of the legacy solution
2. Conducted an in-depth research at the shooting range
3. Proposed a promising hypothesis, considering the defined goals
4. Composed a successful POC
In the course of further cooperation, our team can cover:
1. UI/UX design
2. The reinterpretation of the main features by leveraging computer vision along with sound detection – the concept can be applied either as a full-fledged refactoring of the solution’s functionality or a separate mode, which might be enabled on request
Tech stack:
- Android SDK
- Flutter
- OpenCV
Timeline:
- August 2022 – February 2023
Team:
- 1 BA specialist
- 1 CV engineer
- 1 mobile developer
- 1 UI/UX designer
Business value
The conducted discovery phase is building the foundation for the future product and overall business efficiency. By providing outside knowledge and experience, we evaluated the solution’s strongest and weakest points explicitly highlighting the issues to resolve and opportunities to seize.
Our client can now profit from:
- Independent perspective
- Technical expertise
- A problem-solving approach
- And valuable insight
In the course of further cooperation, we see promising opportunities for both:
- The end-user
– Automatic tracking for greater time savings and accuracy
– Timeline playbacks to track shooting progress
– Event editing to change historical records
– Event filtering to ignore external factors (weather conditions, target movement, and other) - The business
– Higher competitiveness – at the very moment, there are no vendors on the ammunition market that provide ready-made solutions based on computer vision enhanced with sound detection
– Sales increase – a smooth user experience along with new features will attract more customers and boost sales accordingly
What’s next?
Speaking about artificial intelligence, we see great potential in the further evolution of the suggested concept. By leveraging advanced technology, we can help users indicate, record and verify shooting events with even greater accuracy eliminating resource-intense manual routines.